University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Chalmers University of Technology | 3.2 Slope Stability
The slope of the dam will be exposed to various pressures when the reservoir level is
frequently changed. What material properties will govern the stability and conditions
at which the slope is stable will be analysed.
Results are presented according to the four scenarios stated in chapter 2.3. Each
scenario has been tested according to stated variations in material properties: friction
angle based on recommendations, applied cohesion and friction angle according to
standards. As earlier stated in the methodology chapter, the freeboard height has been
set to half a meter.
A sensitivity analysis was first done on a three meter high dam, see Figure 15. The
different operating scenarios are presented next to each other with the slope angles 1:2
and 1:2.5. The different bars represent the different material properties, blue for no
cohesion, red with an cohesion of 1 kPA and green for standard properties of friction
angle.
For the normal operating scenarios, the factor of safety reaches the recommended
values. Concerning the extreme scenarios there will be some issues for the slope angle
1:2 when applying a load together with a RDD. An increased cohesion results in a
higher factor of safety and a decreased friction angle causes lower factors of safety.
Figure 15 Factor of Safety for a dam of height three meter according to different
material properties.
For a dam with the height of five meters, there will be an overall decrease in the factor
of safety compared to a dam height of three meters. During normal operating
scenarios there will be no problems with stability. Concerning the extreme cases there
are some problems, especially in case of an RDD with an additional load, see Figure
16.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
22 |
Chalmers University of Technology | Figure 16 Factor of safety for a dam of height five meter, with different material
properties.
The overall result of the stability calculations is that the geometry of the dam affects
the stability. A dam with a height of three meters reaches recommended values for the
factor of safety in almost all scenarios and for the five meter high dam there are some
problems. Lowering the slope angle is favourable for the construction and will
provide a better resistance against failure. In addition, as shown in Figure 15 and
Figure 16, the factor of safety is changed if the material properties are altered.
3.3 Hydraulic Conditions
The hydraulic gradient within the embankment is affected by the variations in the
reservoir level. The investigation will consider which material properties are
governing the internal stability and when there may be a risk of failure.
Calculation show that the hydraulic gradient, determined in a laboratory, will be less
than the evaluated critical gradient. Results are seen in Table 3. The minimum,
maximum and average bars for hydraulic gradient represent the interval of laboratory
results, compared to the critical hydraulic gradient evaluated according to the
minimum densities and porosities. The ratio between critical hydraulic gradients and
the hydraulic gradient are all above two, which indicates that the conditions in the
dam are on the safe side.
Table 3 Comparison between the measured hydraulic gradient and the evaluated
critical hydraulic gradient.
Minimum Maximum Average Unit
Critical Hydraulic gradient 1,26 1,35 1,31 [-]
Hydraulic gradient 0,42 0,52 0,46 [-]
Relation 3,00 2,60 2,85 [-]
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
23 |
Chalmers University of Technology | 3.4 Design Features
According to the present condition of the features and the function, the systems have
been verified by calculations and with consideration to present recommendations.
3.4.1 Design of Dam Crest in Terms of Freezing Depth
Evaluations have concerned the three areas; Stekenjokk, Boliden and Kristineberg.
Expected freezing depths are seen in the Table 4 below, results are displayed for both
a homogenous dam and a zoned dam.
Table 4 The freezing depth in embankment dams in different geographic locations
where mining activity take place within Boliden Mineral AB (SveMin, 2012).
Homogenous Dam Zoned Dam
Z 2,89 3,13 [m]
Boliden
Z 2,99 3,24 [m]
Kristineberg
Z 2,82 3,05 [m]
Stekenjokk
3.4.2 Freeboard Capacity
According to evaluation by Vattenfall (1988) the dimension of the freeboard should
be 0.3 meters. According to ICOLD (2005-2010) the dimension should be at least 0.4
meters. The diverse results may be described due to the difficulty to find balance
between which parameter that will be of greatest importance for dam safety.
3.4.3 Discharge System
To achieve a steady outflow and a steady water level in the basin the discharge system
need to be designed for the specific dam in question. The recommended design has
been chosen to be a V-shaped outlet with a cover of erosion protection material, with
the advantage of using the material found on site. The outflow should be constant and
should not create too much turbulence and thereby affect the sedimentation process.
To avoid risk of erosion damages, an erosion protection material should be applied.
According to the velocity of the water the dimension of the erosion protection
material was set to 0-100 mm. This type of outlet is advantageous when doing visual
inspections. There will still be a risk for erosion damages in the flow path but they
will be easier to inspect. It will also be possible to detect potential erosion in the water
when the water is lead above the construction and not through. Illustration of a section
on the dam crest is presented in Figure 17 below. The reservoir level is rather low and
the flow will be low in relation to the capacity. In the example there is also a metal
plate on top of the discharge, allowing a transport path on the crest.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
24 |
Chalmers University of Technology | 4 Discussion
Providing a conceptual design for a dam for this rather diverse application area will
contribute to some extent of assumptions and generalization. However, the results are
a foundation describing the sensitivity and stability governing aspects, for these
within the mining industry in proportion rather small dam constructions. Hence, the
aspects considered to affect the stability the most seems to be the process and the
operations surrounding the constructions. General evaluations of systems and stability
have been done, there a no site-specific tests. The material investigation applied
represented a specified area and have considered being representative. A
generalization of the material properties was necessary, which should be taken into
consideration when evaluating the results of the thesis.
The function of the dam and the demands on an easily operated and constructed
product has been the main focus for these dams. Small dams operating in the mining
industry are in some considerations exposed to larger loads and stresses of the
material than larger dams with stable reservoir levels. Continuous changes in reservoir
levels will cause a change in the conditions of the material within the dam even during
stable conditions in slopes. Effects on the material behaviour in a long-term
perspective are not investigated in this thesis. However, there is a need to verify if the
material and the material properties will change over time. It is determined that there
is seldom a defined life span for the construction. The dams are built without any
reflection on if they should be operating for 10 years, 20 years or even longer. If an
evaluation for expected effects on the material due to the variations in reservoir levels
is determined there is possibility to see the status of the material and if there is an
effect to the stability of the material.
Simulations concerning extreme conditions have been chosen to be a rapid drawdown
of reservoir level. There is also another aspect, such as external conditions or weather
conditions that would have been of interest to investigate. However, in this study the
majority of the constructions have been considered not to be effect of extreme
precipitation or surface runoff due, causing additional inflow to the reservoir. Also
that it is not common with seismic actions. However, it is necessary to define
additional aspects that may affect dam safety, such as external or site related
consequences.
4.1 Sensitivity Regarding Material Properties
The intention was to investigate whether the differences in the material properties
should affect the stability also for small dam constructions. Results from the
investigation show that the slope stability will be affected from changes in the
material properties however; the geometry will also be of importance. Even if there
have been some stability failures in the area of investigation the main causes have not
been found to be slope failures. This is why the material was assumed to have some
cohesion, which would benefit the slope stability.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
27 |
Chalmers University of Technology | Some cohesion has been proven to have advantages for the slope stability, but an
increase of cohesion may also contribute to a decrease of the critical hydraulic
gradient. The risk for internal instability might be higher in a fine grained material. To
contain stability for both slope stability and gradients within the construction, the
critical properties for both parameters must be evaluated and taken into consideration
when designing the dam.
Compaction of material will be of essential importance for the final properties of the
material. In regulation and guidance for dam construction there are suggestions and
directives for the process of testing results and follow ups on compaction may be
somewhat hard to achieve in a practical manner. This is due to the volumes of
construction material is rather small and the time of constructing a dam may in
comparison be fast. There need to be other methods to determine that the compaction
is sufficient, rather than having continuous testing during the construction.
The material properties of till will be affected of freezing actions during winter, in
relation to the dam height the rather large expected freezing depths need to be
considered. Procedure of removing snow from the crest should be considered when
confirming the stability for a long-term perspective for the dam. Due to the
continuously operating of the water treatment system and the non-acceptance
concerning stop in production, the dams need to be accessed during the entire year.
However, the removal of snow from the dam crest should to the utmost be limited, in
order to protect the material from freezing damages.
4.2 Reflection on Appearance of Small Mining Dams
According to the field trip there are few directives concerning production guidance,
operation procedures and inspection routines. Stability problems have been
considered to be caused mainly by lack in experience and procedures concerning
operating of dam systems.
The freeboard height has been determined to cause larger demands on stability when
the reservoir level is increased. It is needed to have normal operation reservoir levels
and also to determine a level where there is risk in dam safety. If the level in the
reservoir start to increase it is possible to decrease the sensitivity and increase the dam
safety, by having discharge systems with variable discharge capacities. Defects in the
primary discharge system may cause rising levels in the reservoir. To provide a
redundancy of the dam system, one solution would be to have a safety discharge
system, an overflow spillway. The level of the overflow should be determined in
concern to when there is a risk for stability problems in the dam slope.
The usage of geomembranes and different liner systems has been diverse for the
constructions. The usage of geomembranes has not been considered for the stability
calculations. Assumptions made have been that the stability of the embankment needs
to fulfil similar stability demands as dams with liner systems in natural materials.
Usage of both geomembranes and natural liner systems will contribute to a larger
control of seepage of residue water. Systems where the water will be drained through
the sediments and into the ground will not be functional with low permeable liner
systems. Other methods for removing water from the sediments need to be
determined. Removal of sediments will be more sensitive to avoid damages of the
liner system. Both to protect the geomembrane and to avoid damages on the soil
layers of natural liner systems causing potential leakage paths.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
28 |
Chalmers University of Technology | It was determined that in some cases there is a need to have waterproof construction.
The usage of natural materials for these kinds of liner systems demands an extensive
system of different layer compositions. There is a need to define when these situations
may occur and to what extent it should be fulfilled. Whether it is according to
legislation demands or company policies it would contribute to transparency within
the decision making.
To find a more unified way of construction small dams within the company there are
in fact clear directives in present regulation such as GruvRIDAS (SveMin, 2012).
Whether the regulation is applicable in all areas could be further investigated but in
some areas it has been determined to be possible. For evaluating material properties
and features concerning safe operation there are possibilities to apply GruvRIDAS
also for small dam constructions. Thought there may be difficulties when caring out
continuously testing during construction, when the production is not as time
consuming as larger dam projects. There need to be other methods to verify that the
final construction will fulfil stated demands and provide a safe dam environment.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
29 |
Chalmers University of Technology | 5 Conclusion
This thesis has been based on a field study in the area of Boliden. All the five mining
sites have been visited to gather information concerning performance, function and
environment of operating dams. To evaluate the condition of the dams and the
performance regarding demands and stability, the general model was stated together
with a conceptual design. The two loading conditions, water load and internal
hydraulic load, were determined to be main parameters affecting the stability.
The material investigation and the verification of the material have led to a conclusion
that the materials in the area are suitable for construction of small dams. It is also
determined that the variation in the material properties cause instability problems for
the dams. The results show that characteristics of the material will contribute
differently to the stability. Materials with cohesion will benefit the slope stability but
will increase the hydraulic gradient, which in turn may cause internal instability.
Problems with stability for the dam of interest have been determined to be caused by
lack of clear processing of the dams; both during construction and while operating the
dams. This needs to be applied before the dams have been taken into operating.
The small dams operating in different water treatment facilities have been seen to be
of low priority despite the importance of dams within the water treatment system.
With this in mind, it is recommended to always have a redundancy in the systems to
decrease the sensitivity in specific details in dam composition. For example,
overtopping and leakage of untreated water or sediments may be avoided if spillways
and seepage collection techniques are applied.
The verification of stability of small dams has been found to have similarities with the
appearances from the field investigations. Constructions found to have stability
problems have in some cases also been found on the sites. Determination and further
investigation of material properties and the geometry of the dams will provide a safer
operation of small dams within the area.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
30 |
Chalmers University of Technology | The extent of geomembrane differs among the sites and the opinion is that in some
cases the usage of geomembrane is dependent on when the water have not yet been
treated. The geomembrane appears to have been avoided when there is a need to
empty the reservoir from sediments. When geomembrane have been used there is a
system with a sludge pump to remove potential residues.
Figure 22 Water storage reservoirs connected with an overflow threshold.
Reservoir with the intention to contain sediments will occasionally be drained to
excavate the material; the water will be percolated through the underground when the
inflow of water is stopped. The process may appear differently due to the composition
of the sediments. Removal may be carried out by excavator that is located on the crest
of the dam and a dumper transporting the residues to a landfill for storage. If the
location of the dam permits a tractor may be driven down in the reservoir provided
that the sediments are stable. One example of a sedimentation basin is seen in the
Figure 23 below. This is rather time consuming procedure and dependent on the
permeability of the underlying foundation.
Figure 23 The sediments have been drained and waiting to be excavated.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
37 |
Chalmers University of Technology | There have been attempts to pump the sediments from a reservoir. By lowering the
water level to somewhat above the surface of the sediments and then pumping the
sediments away with for example a sludge vehicle. The water level may need to be
lowered in several steps to contain the consistency of the sediment in a sufficient
dissolved condition and to avoid pumping too much water.
The geometry of the constructions is as seen different according to application area
and geometry. The heights have been determined to vary between 1-5 meters and the
slopes from 1:5 to 1:2.5. Below in Figure 24 a drawing of a section from one of the
sites is displayed. The construction consists of homogenous material and are
reinforced with rockfill on the downstream slope due to instability and extended
seepage.
Figure 24 Section of a dam, reinforcement with rockfill.
Discharge systems may due to the surroundings appear differently and there have
been several solutions found in the sites. Below in Figure 25 the two basins are
connected with pipes beneath the water surface. The construction consists of a till
embankment with an erosion protection material on the reservoir side. This solution is
rather common according to the field study. Disadvantages seen are though the
limited possibilities to do inspections of the conditions of the pipe and the possible
exposure of internal erosion in connection to the pipe.
Figure 25 Section of two reservoirs operating as a system, connected with a pipe.
For the sedimentation reservoirs there are solutions with overflow dikes, plastic and
concrete pipes through the dam other examples could be thresholds in both natural
material and concrete. The solutions are dependent on the intention with the reservoir
and the elevation of the connecting reservoir. Thresholds in till or concrete located on
top of the crest have advantages in terms of inspections, as well as redundancy for
dam safety if the level in the basin is increasing.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
38 |
Chalmers University of Technology | In Figure 26 a concrete threshold are shown, containing two levels with the possibility
to increase outflow capacity when needed. In times of lower levels the velocity of the
water will be kept high and prevent risk of freezing during winter time.
Figure 26 Concrete threshold in a dike connected from a system of reservoirs.
Reflections carried out from the field study have been that there are problems
concerning the stability. In some cases there has been instability in slopes causing
failures. Some of the sites have been described having high reservoir levels and an
extended seepage have been noted. There are indicated problems concerning seepage
through the constructions and also sinkholes have been found on downstream slopes.
Sinkholes may be indications of process of internal erosion; the intensity of then
erosion may led to stability problems and needs to be considered as a potential safety
risk.
The extensive material investigation by Mattsson (2007) was initiated before the
construction of a large dam for tailing deposits in the area of Boliden. In the
investigation there were 165 test pits and 20 samples were sent to laboratory for grain
size analysis. Five samples were also tested for hydraulic properties. Laboratory test
was carried out according to Nippel permeameter and Universal permeameter.
Conditions determined according to sieving of material have been considered in the
evaluation of hydraulic gradients. The material assumed to be governing has been
silty till and the amount of fines has shown to be diverse.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
39 |
Chalmers University of Technology | Figure 28 Illustration of the dam in GeoSlope2007 and the responding factor of
safety.
The different scenarios, steady state, steady state with load, rapid drawdown and
rapid drawdown with load have been applied for each of the geometries. For the
program to respond for a scenario where a sudden rapid drawdown have occurred you
need to find a point where the effective stress are zero. This is simulated by assuming
steady state conditions on both sides of the construction and then on one side there is
a rapid drawdown of the reservoir. The piezometric line has been added along the
slope, this will display the exact moment where the gradient have not yet started to
decrease within the construction. Responding to the moment where there is no
resisting load on the slope and assumed to be the most critical moment.
In the Figure 29 below the extreme scenario when a rapid drawdown have occurred
on one the right hand side of the dam. The dotted line represents the piezometric line
and the solid line on the crest and at the dam toe indicates the entry and exit point of
the expected failure zone. The expected factor of safety for this specific geometry and
the stated boundary conditions are seen to the right in Figure 29.
Figure 29 Responding factor of safety when applying load and simulating a rapid
drawdown of reservoir.
CHALMERS, Civil and Environmental Engineering, Master’s Thesis 2014:18
41 |
Chalmers University of Technology | Abstract
Abstract
Due to a very competitive market, forcing companies for cutting cost, each waste is important to
deal with, for instance longer lead times than needed or low delivery precision. One of Sandvik
Mining’s lego supplier’s delivery precision and lead times have been varying, which has created
the interest in investigate the relationship between the companies in order to identify
improvements. An investigation in form of a case study was performed by a student group at
Sandvik Mining in order to investigate this relationship that Sandvik Mining has with this lego
supplier, Ockelbo Lego-‐Mek (OLM). The processes, material and information flow between the
companies has been mapped and further analyzed. The analyses are based on interviews,
observations and data provided by Sandvik Mining. The analysis shows that there is a great
potential for improvement in all areas. Therefore, a new material flow is presented, giving
Sandvik Mining the possibility to cut costs in form of holding and transportation costs.
Furthermore, the analysis of the information flow proposes a new organization structure
towards OLM considering standard products while the already existing product organization
structure should be kept when considering new and test products. The new organization
structure ought to enable a more straightforward communication which also should eliminate
some of the existing issues, for instance priority issues. Finally, an analysis regarding the
processes is presented and shows that processes involved with OLM are in lack of control
documents and are in need of standardization in order to enable continuous improvements. The
conclusions are wrapped up and presented as an action plan, however some conclusions has
been questioned by the student group.
Keywords: material flow, information flow, processes, supply chain management, supplier
development, cost savings
ii |
Chalmers University of Technology | Introduction
1 Introduction
In this chapter the case study is introduced. The background of the problem, the reason for the
case study and the purpose with the case study is also presented. All information without
references is referred to the interviews.
1.1 Background
In order to survive and become a strong competitor companies need to get the right products, at
the right price and time. This puts responsibility on the suppliers´ delivery and quality precision.
To achieve this and in order to match supply and demand it is required that uncertainties within
the supply chain is reduced as much as possible. This requires the information flow to be
constant, accurate and in time, which in turn facilitates to create a good product flow (Lambert &
Cooper, 2000). (Christopher, 2001)
Presently a vast majority of all companies outsource a part of their production. One reason is
that companies lack core competencies for certain products and rather spend their time on
processes they do distinctively well. Hence, letting other companies produce a part of their
products. In order to have a good relationship it is necessary to have, good communication
between customer and supplier, mutual benefits, shared goals and realistic expectations from
both parties. Without these elements there is a risk for disappointment and a poor relationship.
Furthermore, outsourcing is linked with more transportation and therefore it is important that
those are efficient in order to keep the transportation costs as low as possible. (Logan, 2000)
Sandvik Mining which is the investigated company in this case study outsources a part of their
production to ordinary supply chain suppliers as well as to one lego supplier Ockelbo Lego-‐Mek
(OLM). The problem is that the relationship between Sandvik Mining and OLM has become on a
friendly basis with no clear boundaries or authorities. In addition, there seems to be more
interfaces towards OLM than needed. As a consequence Sandvik Mining is facing several
problems with the information flow towards OLM. Furthermore, the material flow seems not be
optimal. The delivery precision for instance is low and the lead times are inaccurate and volatile.
In addition, priority issues seems also to be a problem due to all the interfaces towards OLM,
which might have an influence on the lead times and delivery precision. This has created an urge
to investigate the relationship with OLM regarding both the material and information flow in
order to identify possible improvements in these two areas.
1.2 Company background
Sandvik AB was founded in 1862, by Göran Fredrik Göransson in Sandviken. Sandvik AB first
started out by producing steel in a successful way due to that Göransson were the first one who
successfully managed to use the Bessermer-‐method in an industrial scale. Sandvik AB has during
the years had different markets such as production of saws which 1999 was phased out. Sandvik
AB’s strategy is “One Sandvik to be number one”, with the ambition to be number one in every
business area. Sandvik AB is today divided into five business areas, which are Sandvik Mining,
Sandvik Machining Solutions, Sandvik Materials Technology, Sandvik Construction and Sandvik
Venture. The organizational structure is illustrated in Figure 1-‐1 below (Sandvik, 2013).
2 |
Chalmers University of Technology | Introduction
Figure 1-‐1: Organizational structure.
Today is Sandvik AB represented in 130 countries and has about 50 000 employees where 5500
are located at the facilities in Sandviken. The revenue for 2011 was 94 billion SEK and the profit
was about 5,8 billion SEK. (Sandvik, 2011)
1.3 Ockelbo Lego-‐Mek
Ockelbo Lego-‐Mek (OLM) is a lego supplier located approximately 35 kilometers from Sandvik
Mining in Sandviken. The company is 100% dependent on Sandvik Mining and has been so ever
since the company was founded in 1984. OLM has currently 27 employees and a revenue of
32MSEK (Allabolag, 2011). OLM is seen as a natural complement to Sandvik´s production since
they produce small batches of odd products with which Sandvik Mining does not want to
interrupt their main production. The company is flexible in terms of capability to produce a
variety of products and has the reputation of delivering highly qualitative products.
1.4 Problem description
During 2012 the delivery precision from OLM has been varying a lot, with an average of 76 %,
which has resulted in varying lead times. This problem seems not to depend only on OLM, but
also on Sandvik Mining. One of the biggest contributors to this problem seems to be poor
information flow between the two companies, which according to Christopher (2001) is one of
the most important aspects to become a strong competitor. One other aspect that might affect
the delivery precision is the amount of interfaces between the companies. Furthermore, as
stated previously by Lambert & Cooper (2000) a good information flow enables a good material
flow; therefore it is important to investigate this matter since the material flow is one of the
problems that Sandvik Mining is facing with OLM for the moment. Vanpoucke et al (2009)
argues that supply chains with vast information sharing are performing better and as stated
above the information sharing between the companies is poor which contributes to a varying
delivery precision. Therefore, an increase in information sharing could generate an
improvement in delivery precision. This raises the interest and gives the opportunity to
investigate the issue and thereby identify improvements for the information and material flow
and possibly identify a potential for cost reduction.
1.5 Purpose and research objectives
The purpose with this case study is to clarify Sandvik Mining´s overall relationship with OLM by
mapping the material and information flow between these two companies and also the
processes of the different roles at Sandvik Mining that are involved with OLM. In addition,
improvement proposals and recommendations for these subjects will be given.
To fulfill the purpose, three main research questions with objectives have been identified and
will be investigated. All investigations should include the roles that are involved and briefly
3 |
Chalmers University of Technology | Introduction
describe what is done in each step. To answer the research questions a few objectives have been
developed, which can be seen below
• How does the material flow between Sandvik Mining and OLM look like in the current
situation?
o Map the material flow between Sandvik Mining and OLM
o Pinpoint the amount of incoming material and frequency of the deliveries
o Investigate if there is enough space and resources to handle the incoming
material
• How does the intra and inter information flow between Sandvik Mining and OLM look
like?
o Investigate which interfaces, channels and structure of communication exists
o Investigate which priority rules that are followed regarding the decision of what
to put on lego at OLM
• How does the process of each role involved with OLM look?
o Map how the work procedures looks including the quality insurance and follow
up process
o Investigate which steering group and documents that exist if any, regarding OLM
After answering the research questions, an analysis will be performed on all the findings, with
more focus on the area where the greatest improvement potential is found. The analysis of this
area will be presented as a business case. Finally, an action plan, improvement proposals and
recommendations will be given, generating a higher delivery precision and more accurate lead
times. This will stand as a base for further development in the future of the relationship between
Sandvik Mining and OLM.
1.6 Delimitations
This case study does not consider any production neither at Sandvik Mining nor OLM. All
mappings and flow charts consider only the information and material flow regarding OLM. The
depth of the investigation will be adapted with consideration to the size and believed
importance of that specific area, and also to the time restriction of the case study.
4 |
Chalmers University of Technology | Theoretical Framework
2 Theoretical framework
This chapter contains academical theories that are used to support the current situation at
Sandvik Mining. The theories that are used are relevant for the purpose and research objectives
of the case study. The chapter starts with general theories about supply chain management and
continues with some theories about lead times and calculation of holding cost. The chapter is
concluded with general theories about processes.
2.1 Supply Chain Management
Supply Chain Management (SCM) is a term that usually appears linked to logistics (Segerstedt,
2009). The definition of SCM according to (Christopher, 2011, s. 3) is:
“The management of upstream and downstream relationships with suppliers and customers in
order to deliver superior customer value at less cost to the supply chain as a whole”.
Christopher (2011) also means that SCM strives to reduce costs and increase value creation
through integrating and improve the whole supply chain. The cost reduction and value creation
is created through conveying the customers´ specific needs upstream in the value chain through
different information flows. Simultaneously as the company has control of the material flow up
and downstream in the value chain in order to achieve an effective material flow. At the end, this
will lead to achieving a higher service level with less resource consumption. An important aspect
in order to establish a good relationship through the supply chain is to create a win-‐win
situation for all parties involved. (Segerstedt, 2009) In Figure 2-‐1 below, there is an illustration
of a supply chain.
Figure 2-‐1: Supply Chain Management (Lyson & Farrington, 2012, s. 93).
As mentioned above SCM should cooperate with every entity in the supply chain in order to
make a smoother supply chain. Collaboration between all departments in the focal organization
in together with information sharing and usage of SCM processes should be combined in order
to create a well-‐working supply chain. In order to have a successful supply chain four critical
8 |
Chalmers University of Technology | Theoretical Framework
enablers have been identified by Lyson & Farrington (2012). These are listed below with
consideration to their importance, where number one is the most important. (Lyson &
Farrington, 2012)
1. Organizational infrastructure
2. Technology
3. Strategic alliances
4. Human resource management
(Lyson & Farrington, 2012, s. 95)
2.1.1 Supply Chain enablers
As seen, in order to have a well working supply chain it is important to have an organizational
structure that enables collaboration with other organizations. Important attributes of an
organizational structure include:
• Having a coherent business strategy that aligns business units towards the same goal
• Having a formal process – flow methodologies to enable SCM improvements
• Having the right process metrics to guide the performance of operating units towards
the strategic organizational SCM objectives
(Lyson & Farrington, 2012, s. 95)
Technology is the second most important enabler, it is important to consider how intercompany
relationships are build. Important attributes of technology include:
• Having operations, marketing and logistics data coordinated within the company
• Having data readily available to managers and the coordination of operations, marketing
and logistics data between supply chain members.
(Lyson & Farrington, 2012, s. 95)
The third most important enabler is the strategic selection of allies in the supply chain and in
order to make this work it is important to have the following attributes:
• Having expectations clearly stated, understood and agreed to upfront
• Collaboration on supply chain design and product and service strategies
• Having top management of partnering companies interface on a regular basis
• Having compatible IT systems.
(Lyson & Farrington, 2012, s. 95)
The fourth and last enabler for a well-‐functioning SCM is the human resource management,
important attributes of this aspect include:
• Sourcing, hiring and selecting skilled people at all management levels
• Finding change agents to manage SCM implementation
• Having compensation and incentives programs in plan for SCM performance
• Finding internal process facilitators knowledgeable about SCM
(Lyson & Farrington, 2012, s. 96)
9 |
Chalmers University of Technology | Theoretical Framework
2.1.2 Organizational structures
The structure of the organization has everything to do with execution, the way of how the
organization is build up is in order to match and fulfill the strategy that is set for a company.
From a small and simple functional structure to a large and complex matrix organization, how
are the operating units organized in order to achieve customer value? (Carr & Nanni Jr, 2009)
Galbraith, Downey & Kates (2001) also state that organization structure is a vital part of the
organizational design, in order to achieve the strategy and aim of a company. The design of the
structure is a key success factor, not only on a divisional level but also on an individual level.
That is why it is important to define the responsibility and authority for every involved part in
the organization. There are different kinds of ways to organize the structure of a company. The
structure can e.g. be based on functional, product, customer and front-‐back hybrid. These are
further described below (Galbraith, Downey, & Kates, 2001)
2.1.2.1 Functional organization structure
A functional structure is based on organizing around major activity groups such as operations,
research and development (R&D), marketing, finance and human resources (HR). The company
is divided by the function of every division and the advantage of the structure is knowledge
sharing, specialization, leverage with vendors, economies of scale and standardization while the
disadvantages are the lack managing of diverse product or services and lack of cross-‐functional
processes. (Galbraith, Downey, & Kates, 2001).
2.1.2.2 Product organization structure
A product division is where a company is divided by the product it is producing. Each division
has its own functional structure to support its product/products. A product structure often
evolves from a functional structure when a company grows and diversifies its product or service
lines, when these lines become large enough to support their own production. The advantages
of this structure are, more rapid product development cycles, products are developed to
excellence and there is a broad operating freedom. While the disadvantages are divergence due
to that divisions work independently, duplication of resources, loss of economies of scale and
multiple customer points of contact. (Galbraith, Downey, & Kates, 2001)
2.1.2.3 Customer organization structure
A customer structure divides the company around major markets segments such as industries,
customer groups or population groups. While functional and product organizational structure
have internal advantages, customer organization structure is more based on the customer to
make it easier for the buyer to do business with the organization. The advantages of this
structure are customization of products or services, relationships with customers and the
possibility to offer solutions. While the disadvantages are the same as in product organization
structure i.e. divergence, duplication and loss of economy of scale. (Galbraith, Downey, & Kates,
2001)
2.1.2.4 Front-‐back hybrid structure
The front-‐back hybrid structure combines the elements of both product – and customer
structures in order to gain benefits from both. It allows for product excellence in the back end of
the company in combination with increasing customer satisfaction at the front end of the
company. The advantages of this structure are single point of interface for customers, cross-‐
selling, value added systems and solutions, product focus and multiple distribution channels.
While the disadvantages are contention over resources, disagreements over price and customer
needs, determining the placement of marketing, conflicting metrics and information and
accounting complexity. (Galbraith, Downey, & Kates, 2001)
2.1.3 Information flow in the Supply Chain
As mentioned earlier, technology is one of the most important enabler for a successful supply
chain i.e. sharing information both internally and externally, making information available for
10 |
Chalmers University of Technology | Theoretical Framework
other companies that are part of the focal organization´s supply chain (Lyson & Farrington,
2012). Due to that technology evolves, organizations tend to integrate more. Therefore,
information sharing has become critical when improving the performance of the supply chain.
(Zhou & Benton, 2007)
There is a relation between the amount of information shared and the overall performance of
the supply chain. Supply chains with less information sharing perform poorer in comparison to
supply chains that use more information sharing. Inter-‐firm information flow is an important
factor of supply chain management. Potential benefits of information sharing might be supply
chain coordination and decreased supply chain cost (Vanpoucke, Boyer, & Vereecke, 2009).
2.1.3.1 Collaborative planning, forecasting and replenishment (CPFR)
CPFR is a collaboration process between organizations in the supply chain, whereby they can
jointly plan different key supply chain activities with the aspect of the whole supply chain, from
raw material to customer. (Blackstone & Cox, 2005)
CPFR is a web-‐based attempt to coordinate various activities such as, production & purchasing
planning, demand forecasting and inventory replenishment between different organizations in
the supply chain. The aim of CPFR is to exchange selected information on a shared web server in
order to provide reliable information and long term future views of demand between the
organizations in the supply chain. (Fliedner, 2003)
As Fliedner (2003, p.16) states “The potential benefits of sharing information for enhanced
visibility in the supply chain are enormous”. The potential benefits of using CPFR differ
depending what kind of actor the organization is in the supply chain. (Fliedner, 2003)
• Retailer benefits
o Increased sales
o Higher service levels
o Faster order response times
o Lower product inventories, obsolescence, deterioration
• Manufacturer benefits
o Increased sales
o Higher order fill rates
o Lower product inventories
o Faster cycle times
o Reduced capacity requirements
• Shared supply chain benefits
o Direct material flows (reduced number of stocking points)
o Improved forecast accuracy
o Lower system expenses
(Fliedner, 2003, s. 17)
2.2 Lean Production
Lean Production has its roots in the Japanese automotive industry, beginning in the 1950s. To
keep it simple the basic idea with Lean Production is to reduce the time between customer order
and delivery, by eliminating waste (Liker & Meier, 2006). This initiative resulted in Toyota
Production System (TPS) which today is synonymous with Lean Production. (Segerstedt, 2009)
Lean Production means that a company´s resources are used efficiently and that no excess in
resources are used in order to produce efficiently. Hence the purpose of Lean Production is to
identify and eliminate all the activities that does not add any value to the product, in other words
identify and eliminate all waste. (Olhager, 2000)
11 |
Chalmers University of Technology | Theoretical Framework
2.2.1 Waste
According to Liker & Meier (2006) there are eight different kinds of wastes that a company
should eliminate in order to become a leaner company and survive in the tough market. They
also mean that every process whether it is a business or manufacturing process has waste,
regardless if it is a production line process, order taking process or a product development
process. The different kinds of wastes are described below. (Liker & Meier, 2006)
• Overproduction – When producing too early or in greater quantities than the customer
need. This in turn generates other wastes such as overstaffing, transportation cost and
excess inventory for instance.
• Waiting time – Staff watching a machine or waiting for the next processing step, due to
no stock, capacity bottlenecks or equipment downtime for instance.
• Transportation or conveyance – All type of movement of work in progress up and
down in a process. This also includes moving material or finished goods from or to
storage between processes.
• Over processing or incorrect processing – When producing products with higher
quality than is necessary, when steps unneeded are taken to produce a product or when
processing inefficiently, due to poor tool or product design which results in producing
defects.
• Excess inventory – All work in progress, finished goods or raw material that causes
obsolescence, transportation and storage costs, longer lead times or delays.
Furthermore, excess in inventory might hide late deliveries from suppliers, production
imbalances, defects, set up times and equipment downtime.
• Unnecessary movement – All movement including walking that is not value adding for
the product e.g. looking for, reaching for or stacking parts.
• Defects – All defective products or products that need to be corrected including repair,
scrap and additional production.
• Unused employee creativity – All ideas, skills improvement possibilities that are lost
by not engaging or listening to the employees.
(Liker & Meier, 2006)
2.2.2 Standardized work
Standardization is about performing a task according to the currently best known established
solution. The work performance is only to be changed when a better solution is identified.
(Segerstedt, 2009)
Further Liker & Meier (2006, s. 124) mean that standardized work is a prerequisite for
improvements by stating the following.
“If the work is not standardized and it is different each time, there is no basis for evaluation”
This means that if no standardization is settled there is no reference point from which to
compare. Therefore, it is important to have a standardized way of performing the tasks so that
improvement can be made from a reference point with the currently best known solution. One of
the main prerequisites for having a standardized work is that the work is repeatable. (Liker &
Meier, 2006)
2.2.3 Lead time
Lead time is a concept of time that can be used in different situations. The general definition for
lead time is the time it takes for one part to make its way through manufacturing, beginning with
arrival as raw material to shipment to the customer. (Rother & Shook, 1999)
12 |
Chalmers University of Technology | Theoretical Framework
Olhager (2000) argues that lead time can be seen from different perspectives which are for
instance, the lead time it takes to develop a new product, the lead time for delivery in customer’s
perspective and also the lead time to customer in the producing company’s perspective.
Lead time is also connected to tied up capital such as inventory. A reduction in inventory lowers
the lead time meaning that products reach the market faster when inventory is lowered.
(Srinivasan, 2004)
2.2.4 Tied up capital
A company’s assets can be divided into fixed assets and turnover assets. All of the assets have a
monetary value and examples of fixed assets can be land, buildings and machinery while
turnover assets can be inventory, transportation and production cost. (Jonsson & Mattsson,
2005)
When doing investments, capital is tied up and which affect the company’s cash flow while it also
generates a cost, corresponding the income the money would have generated if they were e.g.
put on a bank account (Jonsson & Mattsson, 2005).
Tied up capital affect a company’s profitability directly and also the delivery service indirectly
i.e. if the inventory would be lowered too much it would result in bad delivery performance. The
average tied up capital indicates how much money is tied up in inventories, work in progress,
finished stock and transportations. The tied up capital can be presented in absolute numbers,
but if this is not possible, it can also be present as inventory turnover rate and average lay time
of product in inventory. (Jonsson & Mattsson, 2005)
2.2.5 Holding cost
All inventory that is not tied to a specific customer order runs the risk to not be sold, hence
increasing the holding cost (Gudehus & Kotzab, 2012). The holding cost is the amount of money
a company has to pay to keep material in stock. The holding cost includes warehousing,
obsolescence, pilferage, damage, insurance and taxes. (Timme, 2003).
Jonsson & Matsson (2005) sums all of these variables into three and presents a formula of how
to calculate the holding cost interest. The formula is presented below:
Holding cost interest =
capital avoidable cost+warehouse avoidable cost+contingency avoidable cost
𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑣𝑎𝑙𝑢𝑒 𝑠𝑡𝑜𝑐𝑘
All of these variables that are included in the holding cost interest which is presented as a
percentage number, so when calculating what the holding cost for an inventory is this
percentage cost is multiplied with the average value of the inventory during a year (Jonsson &
Mattsson, 2005). Example, average value stock is 1 350 000 SEK and the holding cost interest is
15 % meaning that the holding cost for this specific inventory is 202 500 SEK, the calculation is
presented below.
0,15∗1 350 000 𝑆𝐸𝐾 = 202 500 𝑆𝐸𝐾
2.3 Delivery service parameters
The service considering the accomplishment of order-‐to-‐delivery process is often mentioned as
delivery service. This process includes the phases from order until delivery and during the
delivery itself. To explain delivery service there are a couple of delivery service parameters used
in order to describe the delivery performance. The importance of the parameters varies
13 |
Chalmers University of Technology | Theoretical Framework
depending on the situation it is describing. Following is a description of the most used delivery
service parameters (Jonsson & Mattsson, 2005)
2.3.1 Delivery precision
Delivery precision explains to what extent deliveries arrive at right time i.e. the time that the
customer and supplier have agreed on. Delivery precision differs from warehouse service level
in that matter that delivery precision considers only articles that are not in stock but articles
that have to be assembled or produced directly to order (Jonsson & Mattsson, 2005). Delivery
precision can be applied both externally and internally, between departments, in a company.
(Madhusudhana Rao, Prahlada Rao, & Muniswamy, 2011)
Delivery precision can be measured as the ratio between delivered orders on time and in
comparison to total number of orders. The delivery point can be a single day or an interval of
days and this is something that is agreed between the supplier and customer depending on the
product itself and the demand of the product. (Jonsson & Mattsson, 2005)
2.3.2 Delivery assurance
Delivery assurance measures the deliveries quality in terms of if it is the right product being
delivered and if the quantity is correct. When having a low delivery assurance it often leads to
unnecessary activities, which would not occur if the delivery assurance would be satisfying.
Jonsson & Mattsson (2005) states that delivery assurance can be measured as the ratio between
the number of orders with remarks (wrong product or wrong quantity delivered) in comparison
to the total number of orders sent. (Segerstedt, 2009)
2.3.3 Delivery time
Delivery time is the time it takes from the point an order is received until products are delivered.
Delivery time consists of administrations and order processing time, dispatch and
transportation time and in some cases design and manufacturing time. Delivery time is normally
expressed in days or weeks. The longer delivery time, the poorer flexibility due to that orders
take longer time to deliver. This results in an increase in tied up capital since material is tied for
a longer time. (Jonsson & Mattsson, 2005)
2.3.4 Delivery flexibility
Delivery flexibility considers the capability to adjust to change in customer demand. Changes
could be in time, quantity or even changes in products themselves. There is a difference in
delivery flexibility before received order and during a received order. Delivery flexibility before
received order concerns the possibility to accept changes in delivery time, minor order
quantities than agreed or changes on products. While delivery flexibility during a received order
concerns the possibility to adjust to higher demand in short time and to changes such as to
change delivery date on orders or deliver higher quantity than agreed. (Jonsson & Mattsson,
2005)
2.4 Processes
What is a process? According to Bergman & Klefsjö (2010, s. 456) “a process is a network of
activities that are repeated in time, whose objective is to create value to external or internal
customers”. Due to that there are a lot of different activities that can be called a process, a
classification of processes has been done. Processes in an organization has been divided into
three groups (main, support and management processes) which are illustrated in Figure 2-‐2
below.
14 |
Chalmers University of Technology | Theoretical Framework
Figure 2-‐2: Processes in an organization (Bergman & Klefsjö, 2010, s. 458).
• Main processes -‐ These processes’ task is to fulfill the needs of the external customer and
refine the products that are provided to the process. These kinds of processes are in a
way “the life nerves” of the organization since the processes’ output is what generates
the income for the organization. Examples of this type of processes are product
development processes, production processes and distribution processes.
• Support processes – These processes’ task is to provide resources for the main processes
and most often these processes have internal customers. Examples of this type of
processes are recruitment, maintenance and information processes
• Management processes – These processes’ task is to make decisions regarding the targets
and strategies of the organization, and to implement improvements into other
organizational processes. Likewise support processes the management processes most
often have internal customers. Examples of processes are strategic planning, targeting
and auditing
(Bergman & Klefsjö, 2010, s. 458)
2.4.1 Process flow analysis
Process flow analysis is a method used to document activities in detail and graphically as basic
data in order to give a better understanding of the process and clarify potential process
improvements. A process flow analysis can be performed on all three types of processes
mentioned above. Different types of schedules and charts are preferably used to describe and
analyze processes and organizations. The analyses that are made with the charts may have
different purposes, hence why a process flow analysis can vary in level of detail and information.
It can concern a production process in its fullness including all the activities, a part of a
production process, or a detailed mapping of individual processes. When doing the actual chart
different types of symbols are used for different activities. (Olhager, 2000) The fundamental
steps in a process flow analysis are the following:
1. Identify and categorize the process activities
2. Document the process as a whole
15 |
Chalmers University of Technology | Theoretical Framework
3. Analyze the process and identify possible improvements
4. Recommend appropriate process changes
5. Perform decided changes
(Olhager, 2000, s. 92)
When analyzing the process each work activity in the process chart is scrutinized through the
questions What? When? By who? Where? For how long? How? and especially Why?. Why is this
process performed at all? Why is it done in this way? Other questions might be when, where and
how could it be done differently? (Olhager, 2000)
2.4.1.1 Block diagram
Different kind of tools can be used when observing a process and one way to do this is by using a
type of mapping called block diagram (Bergman & Klefsjö, 2010). Blackstone & Cox (2005, s. 11)
describes block diagram as “A diagram that shows the operations, interrelationships, and
interdependencies of components in a system.” Block diagram may also be referred to as
flowchart or process flow chart (Blackstone & Cox, 2005). Figure 2-‐3 below illustrates a
flowchart.
Example: Production process
Company A Company B
Start
Process
Process
Decision Process
Process
Document
Process
End
Figure 2-‐3: Example of a flowchart.
16
ssecorp
noitcudorP
:elpmaxE |
Chalmers University of Technology | Method
3 Method
The content of this chapter presents the approach of the case study and which kind of
information that has been gathered and used.
3.1 Type of study
There are different approaches of collecting data for a research project. The approaches for
collecting data can be quantitative or qualitative. If the purpose of the research project is to
collect numerical data, statistics, standardization and generalization then it is recommended to
use a quantitative method. This data can be gathered with polls and inquiry sheets that consist
of questions and data that cohere to the research project. (Olsson & Sörensen, 2011)
On the other hand if the situation of the research project is unique, complex and/ or based on
individual perception then a more qualitative method is recommended. The purpose with
qualitative methods is to characterize a specific task by using models, description or
categorization in order to describe a specific phenomenon. The information for qualitative
studies can be gathered through interviews, observations and/ or literature. When doing
research about a specific case, person, group or social entities, a common term used for this
research method is case study. The gathered information is further used to present a the current
situation and also to do an analysis. The information that is gathered is finally summarized with
a discussion and conclusion. (Olsson & Sörensen, 2011)
In order to fulfill the purpose the authors decided to perform a case study in order to get an
understanding of how the relationship between Sandvik Mining and OLM looks. To get a deeper
understanding about the relationship, the case study was divided into three main areas, which
are the, material flow, information flow and processes.
In order to get information about how the material flow, information flow and the processes
between Sandvik Mining and OLM looks, the authors chose to use flow charts. A particular
flowchart used is called block diagram, which gives the opportunity, besides mapping the
process, to identify where in the organization the process is performed. (Bergman & Klefsjö,
2010).
Most processes have a great potential for improvement, therefore it is often worth the effort
performing these mappings. As Bergman & Klefsjö (2010, s. 462) state, “The knowledge that is
created by defining and mapping a process is highly valuable in itself. In addition, it is an
excellent platform for the improvement work, as it generates a shared picture of current events”.
(Bergman & Klefsjö, 2010)
Every employee that was found necessary for the case study was interviewed. The aim with
these interviews was to sketch the block diagrams and get the most truthful picture regarding
the relationship and interfaces between Sandvik Mining and OLM. Furthermore, the aim with the
block diagrams in this case study was to illustrate the work-‐processes of each role involved with
OLM, which is one of the research questions in the purpose chapter.
By using this method it gave the opportunity to get an overview of how the material and
information flow looks and which processes that are performed in the existing interfaces
between Sandvik Mining and OLM, which is one of the research objectives.
After the mapping, it was clear that all investigated areas had potential for improvements and
was therefore further investigated. One of the areas, the material flow, was identified to have a
higher grade of potential for improvements and was therefore in collaboration with Sandvik
Mining chosen to put most focus on. A business case including cost savings and suggestions for
improvement was made. Due to the deeper investigation in this area, the recommendations are
more thorough than for the other areas´.
20 |
Chalmers University of Technology | Method
3.2 Purpose of method
The purpose of the method is to function as guidance and help for the authors in order to fulfill
the purpose of the case study. When the mapping of the processes was done the flow charts
were the basis for identification of improvements for the material and information flow and
processes between Sandvik Mining and OLM.
3.3 Data gathering
According to (Yin, 2007) there are different kinds of sources of information and these are
presented below.
• Documents
• Interviews
• Direct observation
In combination with these sources the authors has additionally used one more source of
information that is literature. Literature comprises books, articles and homepages and has been
used to gather the necessary literature
3.3.1 Documents
Yin (2007) describes documents as internal documents at a company, which can be used in
order to accomplish the purpose of a case study. Patel & Davidson (2003) divides documents
into different sub-‐groups such as statistical, public, private, figure-‐documents and audio-‐
documents. The authors have used some of these documents, provided by Sandvik Mining in
order to do the case study.
3.3.2 Interviews
When doing case studies interviews can be a very important source of information, information
that only specific persons have and are not on paper. There are two different aspects to be
considered when using interviews, the aspects of standardization and structure. Interviews with
high level of standardization consist of questions that are made up before and used on all
interview objects, while interviews with low level gives the opportunity to make up questions
during the interview. (Patel & Davidson, 2003)
The aspect of structure regards how specific the question is and how much room that is left for
the interviewee to interpret the question. With high level of structure the questions are very
strict and spot on while interviews with low level of structure gives the interviewee the room to
interpret the question in their own kind of way. (Patel & Davidson, 2003) The interviews that
were held by the authors had a quite high aspect of standardization and also a quite high level of
structure. However during the interviews other complementary questions related to the main
questions were added and the interviewees was allowed to give their own input regarding the
questions, which makes the type of interviews held to semi-‐structured.
3.3.3 Direct observation
Observations are one of the best forms when it comes to get information, and by being present
and doing observations. Information can be gathered to see how it really looks at a specific
situation. (Patel & Davidson, 2003) The authors performed some direct observation at the site of
the company where the case study was performed. Observations were performed when the
existing data were not sufficient for the purpose.
3.4 Literature
The authors gathered information through literature in form of books and articles. The authors,
which used books and articles that had been used in previous courses did the selection of
literature. The authors did also use new literature in form of books and articles that was found
necessary for the case study that had not been used by the authors before. The search engine at
Chalmers University Library (SUMMON) has been used as the primary source of information
21 |
Chalmers University of Technology | Method
regarding articles. Besides books and articles, the Internet and supervisor was used to find
literature that was relevant to achieve the purpose.
3.5 Method analysis
When the material and information flow and processes were mapped, the next step was to
analyze these areas. In Figure 3-‐1 the work process and methods used by the authors are
illustrated. Literature that was considered suitable for the case study was gathered through
books, articles, Internet and supervisor. Remaining information, that was considered essential,
was gathered from interviews, observation and archive documents provided by Sandvik Mining.
Figure 3-‐1: Methodological approach
The case study was performed at Sandvik Mining’s site in Sandviken, where regular steer-‐group
meetings occurred. The progress of the case study where controlled by weekly check-‐up meeting
where the progress was presented and discussed with concerned personnel in order to ensure
that the project were heading in the right direction.
3.6 Method discussion
Due to that the purpose of this research project was to map and present the processes, the
material and information flow between Sandvik Mining and OLM, the method of performing a
case study at the site of Sandvik Mining seems to us as a valid approach to this research project,
because it gave us the possibility to create a theoretical paper of a complex reality.
We chose to use block diagram as a tool to start with, which we also find as a good choice not
only because the tool show the process that is mapped but also where in the company this
process is done giving us the possibility to identify the connections between the different roles
and processes.
What we lacked in our opinion when using this tools was that we did not proceed deeper in the
processes, due to that the company wanted us to focus more on the material and information
flow and not the processes themselves. We could have investigated the processes more
thoroughly, identifying waste within every process and the system as whole and made it more
efficient, but as mentioned due to requests from Sandvik Mining and the lack of time this was not
done.
22 |
Chalmers University of Technology | Method
We used different kind of approaches when collecting information. Firstly we had meetings with
our supervisor at the company and went trough all individuals that are involved with OLM in
any kind of way. Later we interviewed these persons where we used a questionnaire we made
with questions that were align with the purpose. The supervisor checked this questionnaire
before being sent to the interviewees, and thereafter the interviews were held. We are of the
opinion that the collection of that was performed in a good way since we collected data at the
spot from persons that are related to the project and also since these questions were checked
before being used, which assures that the questions being asked were relevant and valid.
In order to increase the validity of the flowchart a second interview was held with every
interviewee to confirm that the flowchart that was sketched is correct. One aspect that could
have been done better regarding the interviews that could increase the validity was to interview
all of the individuals that are involved with OLM, but already from the beginning the amount of
interviewees was limited due to that many of them had similar work tasks.
The other part of the information that was gathered was received from the company in form of
documents or direct observation that were performed by the authors. The information from the
company is considered as valid, due to that this information is the same information as the
company is using but also because it is the only information available. However the observations
by the authors was only done once, but in order to increase the validity of the observation it
could have been done several times.
Due to that we have used a known tool when mapping the processes and interviewed employees
involved directly, the reliability of the research study is high. We are of the opinion that the
reliability of the process maps is high because they should be the same regardless of who creates
them. We are of the same opinion regarding the information and material flow due to that both
are based on information received from the company or what we have gathered. The issue that
can be questioned regarding the reliability is if a different research group would have done this
research, the interview questions would probably have been different and thereby other
important information would have been gathered. This could have made the research group to
take a different approach to the task but due to the purpose we still believe that even though
different approaches would have been used, similar result would have been achieved.
Furthermore, during the whole case study, steering meetings were held with the authors,
supervisor and manager at Sandvik Mining. This increases the reliability of the case study since
the supervisor and manager have been updated and given the possibility to comment and
influence on the progress and findings of the case study.
23 |
Chalmers University of Technology | Empirical Data
4 Empirical data
This chapter comprises an overview of the current situation at Sandvik Mining regarding
Ockelbo Lego-‐Mek (OLM), based on personal interviews of employees that are involved with
OLM. All the research questions and objectives are answered in this chapter and a part of it is
further analyzed in the next chapter. The recommendations and improvement proposals are
partly based on this chapter and the following one.
This chapter starts with a description of Sandvik Mining and thereafter continues with
description of the material flow. Furthermore, there are figures in this chapter illustrating the
information flow between different departments and roles at Sandvik Mining that are involved
with OLM. Additionally flow charts describing the work processes of the different roles are
presented. Only suppliers, departments, persons and flows that are involved with Sandvik
Mining´s relationship with OLM are included in these flow charts and figures. Finally, there is a
description of the quality process that occurs between Sandvik Mining and OLM.
4.1 Sandvik Mining
As mentioned earlier Sandvik AB is divided into five business areas and one of them is Sandvik
Mining, see Figure 1-‐1. The case study was performed at the site in Sandviken, Sweden. The
subdivision that today is Sandvik Mining started in 1907 when Sandvik AB started to produce
hollow steel drills (Sandvik, 2011)
Sandvik Mining with its headquarters in Amsterdam, Netherlands, is the second biggest division
considering both number of employees and revenue. Sandvik Mining’s revenue was 32 232
MSEK and the number of employees was 13 300 in 2011. (Sandvik, 2011).
Sandvik Mining is a global supplier of equipment, tools, service and technical solutions for the
mining industry and is producing highly specialized performance products, solutions and
services. Sandvik Mining has ten business segments and these are:
• Rock Tool and Systems
• Drill rigs and rock drills
• Load and haul equipment
• Continuous mining and tunneling machines
• Crushers and screeners
• Conveyor components
• Bulk materials and handling equipment
• Breakers and demolition tools
• Mine automation systems
• Safety and environmental products
(Sandvik, 2011)
4.1.1 Sandvik Mining Rock Tools
Rock Tools is one of ten business segments within Sandvik Mining with focus on manufacturing
of tools used when mining. Sandvik Mining Rock Tools (Sandvik Mining) offer the widest range
of tools and accessories for exploration, rock drilling, raise boring, coal and mineral cutting,
tunneling, trenching, road grading and cold planning. (Sandvik, 2011)
Sandvik Mining has as mentioned one of the widest product segments and in Figure 4-‐1 below
some of the products are illustrated. The production at Sandvik Mining is divided into two flows,
short respectively long products. The short product flow includes products such as; bits (1),
adapters (4), sleeves (5) and thread ends. The long product flow considers pipes (2) and bars
(3).
26 |
Chalmers University of Technology | Empirical Data
OLM OLM
Tempo
Tempo
1
2
Supplier Sandvik Mining Sandvik Mining Sandvik Mining
5 days 16 days
Figure 4-‐2: Current material flow.
Another figure demonstrating the amount of material delivered by each supplier is illustrated in
Figure 4-‐3.
The information sharing between Sandvik Mining and OLM is done both electronically and
physically. The type of information sharing differs depending on the stage of the product. At
Sandvik Mining there are three different stages of product and the first one is when a product is
a test-‐product. At this stage a few number of the product is produced in order to test the product
and see if it is good enough to be taken to the next stage, which is a new product. At this stage
the product is introduced and if it sells good enough it will be taken to the last stage, which is a
standard product. The new and test products have the most uncertain production lead time and
are also the ones with the longest production lead time. However these products only
correspond for approximately 2% of all the orders sent to OLM.
The ordering towards OLM for standard products going through Tempo 1 is made electronically
by EDI and is generated automatically when the material are registered in the ERP after being
received from the raw material suppliers. Standard products that are going to Tempo 2, together
with new and test products that are going either to Tempo 1 or 2 are sent to OLM with a physical
order sheet that is written manually by the goods receiver. Standard products that are sent to
Tempo 2 and new and test products are not registered in the ERP-‐system, hence the reason why
the goods receiver writes manual orders. This is illustrated with the two dark arrows that go
from Sandvik Mining to OLM in Figure 4-‐3. Note that only material sent to OLM is included in
this figure.
28 |
Chalmers University of Technology | Empirical Data
Sandvik
Mining
Purchasing R&D Production
Tactical Goods Order Production
Planning Design Quality
purchasing receiving processing technique
Figure 4-‐4: Organization structure involved with OLM.
As can be seen in Figure 4-‐4 there are three main departments and seven different sub-‐
departments that are involved with OLM. Most of these sub-‐departments have several
employees that are involved and spend different amount of time on communication with OLM,
which affects the usage of OLM´s resources to different degrees. This gets complex at OLM since
there is only one person that is contacted at OLM, the founder of the company. This person has
the role of CEO, purchaser, seller, production manager, production technician, order processor
and planner.
Most of the employees at Sandvik Mining involved with OLM are in need of contacting someone
for questions regarding e.g. drawing, production or quality. Due to the fact that Sandvik Mining
does not have a clear communication structure towards OLM (which answers a part of the first
objective under the second research question in the purpose chapter), the easiest way of getting
the needed information is by contacting the same person (the CEO) at OLM directly, since this
person has several roles.
In Figure 4-‐5 one can see how the information flow goes between the sub-‐departments and
OLM. Note that almost all sub-‐departments have several persons that are involved with OLM, but
this does not mean that all communication looks the same for everybody towards OLM within
this sub-‐department. Some arrows only indicate for certain persons and other arrows indicate
for everybody.
In the middle of Figure 4-‐5 one can see OLM, and the seven different sub-‐departments around,
that contacts OLM for questions regarding products and drawings for instance. OLM does also
contact the different sub-‐departments when something has to be clarified or if any doubt arises.
This means that almost all communication occurs in both ways except for a few ones that only
occur on a single direction e.g. from quality to tactical purchasing.
31 |
Chalmers University of Technology | Empirical Data
Figure 4-‐6: Individual communication network within Sandvik Mining and towards OLM.
Due to that there is no formal structure in the communication flow for the different product
types, the figures Figure 4-‐5 and Figure 4-‐6 illustrates the merged communication flow on a
departmental and individual level for all product types (standard, new and test products). These
figures answers the other part of the first objective under the second research question in the
purpose regarding how the intra and inter information flow towards OLM look like. The CEO at
OLM was interviewed and confirmed that all these roles has some kind of communication with
the CEO.
4.3.2 Prioritization issues and rules
The different persons contacting OLM causes priority problems at OLM according to some of the
interviewees´. For instance when a new product is being developed a lot of time is spent on
communication, setting up machines and testing until an acceptable product is attained. This
might prolong the lead time for standard products with up to four days which causes
prioritization issues at OLM. In some other cases when a designer or order processor calls and
asks how the production of their order is going, it might be interpreted at OLM as a priority call.
This means that OLM might interrupt the production of the current product in order to prioritize
another product, which means that the machines must be set-‐up and thereby the lead times are
extended. This interpretation at OLM is confirmed by several employees at Sandvik Mining even
though OLM denies that priority is given to someone that calls and asks for the status of a
product.
From Sandvik Mining´s point of view there should only be one person that has the authority to
call and prioritize an order. However, as mentioned this is not how it actually works according to
some of the interviewees´ at Sandvik Mining, even though the planner formally is the one with
this authority. From OLM´s point of view there are only two different sub-‐departments (Design
and Planning) that contact them to have their products prioritized, of which OLM prioritizes one
of them more than the other. According to the planner, priority calls are only based on customer
33 |
Chalmers University of Technology | Empirical Data
demand or if the production at Sandvik Mining needs a specific component. There are three
priority rules followed from Sandvik Mining when deciding where production should take place,
which are:
• Available capacity
• Technically possible
• Profitable (frequency & batch order)
The first one considers if there is capacity available at Sandvik Mining. The second considers if
Sandvik Mining has the technical capacity to produce the product and the last one concerns if it
is financially profitable to produce these products with consideration to set-‐up times, frequency
and batch quantity. These rules answer the second objective in the second research question in
the purpose chapter.
In the following sub-‐chapters there are several flow charts, one for each person that was
interviewed with exception of the tactical purchaser, from now on only called purchaser. These
are also results from the same interviews as mentioned earlier. These flow charts were created
in order to map the work processes that each person involved with OLM goes through and to
answer the third research question in the purpose chapter.
4.4 Roles involved with OLM
As mentioned earlier a more thorough investigation of each role involved with OLM were
performed resulting in a block diagram for each role. These are now presented in this sub-‐
chapter.
4.4.1 Designers
There are several designers at Sandvik Mining that are in contact with OLM. Many of them have
similar work tasks when it comes to the procedure towards OLM and therefore only two of them
have been interviewed, which means that the following flow charts are based on these two.
However the two designers contacts different persons for the same type of questions. For
instance one designer contacts the planner for questions regarding the status of a specific order,
while the other designer contacts OLM directly for the same question. OLM does in some cases
contact the designers directly for questions regarding drawings i.e. the communication occurs in
both ways. Also notable is that none of the designers knew anything about any control
documents, which mean that almost everything is based on experience and how it currently is
done.
The whole process for the designers at Sandvik Mining starts with development of new products
that either come from a development project or from a customer order with specific needs. The
development is done as a CAD-‐drawing and sent to the order processors as a test order. The
order processor decides where to produce the new product, and when needed a production
technician is involved in this decision. If the test order is approved without the need of
correction the process for the designers ends, conversely to the case where a correction is
needed. In that case the designer is contacted by a production technician either from Sandvik
Mining or OLM depending on where the product has been decided to be produced. In the latter
case several outcomes are possible e.g. a small or big change might be needed but in some cases
a total redesign is required.
When this is the case the designer corrects the drawing until it is approved and thereafter the
process ends. In the worst case when the product is too complicated to produce, the decision to
cancel the production of this product can be taken. The communication between the designers,
order processors and OLM varies from time to time, in some cases it is more, depending on the
amount of work that is given to OLM at that specific time. Overall the time spent on this
communication is small. The designers’ procedure is illustrated in Figure 4-‐7.
34 |
Chalmers University of Technology | Empirical Data
Designer
Sandvik Mining OLM
1. Test order is sent to order
processor and decision where
Start to produce it is taken. If
needed a production
technician is involved in the
decision.
CAD-‐model
created/
2. Depending on the decision
corrected
of where to produce the
production technician either
at OLM or Sandvik AB
Test order sent contacts designer to inform
to order that correction is required.
processor (1) Several outcomes are
possible, e.g. small or big
changes can be needed or a
total redesign.
Changes
needed?
3. If it occurs that the product
No Yes
is too complicated to
produce, the decision to
cancel production is taken
Production
and thereby the process
technician
ends.
contacts
designer (2)
Too
complicated
? (3)
No
Yes
End
Figure 4-‐7: Designers´ process when developing new products.
In some cases, when the decision to produce at OLM has been taken, even though the drawing
goes through all steps as illustrated in the flow chart above, there could be doubts that need to
be clarified for OLM by e-‐mail or phone. It necessarily do not have to be doubts, it could also be
improvement proposals that has been detected or just a reflection. Nevertheless, in that case a
solution is given to OLM and when no change is needed the process ends, otherwise a change is
done and sent to the order processor at Sandvik Mining for approval. When the new drawing is
approved it is released and the order processor is notified. Most steps are based on experience
and no control documents exist. Figure 4-‐8 illustrates the process when this occurs.
35
rengiseD |
Chalmers University of Technology | Empirical Data
Designer
Sandvik Mining OLM
1. OLM contacts designer for
clarifications, improvement
Start proposals, misses or other
kind of reflections/ doubts.
In some cases it is solved
immedietaly and no further
OLM contacts
actions are needed.
designer for
clarifications (1)
Changed
needed?
No Yes
Change is made
and sent to
order processor
Order
processor
approves new
drawing
New drawing
released and
order processor
notified
End
Figure 4-‐8: Designers´ process when contacted for clarifications.
4.4.2 Order processors
There are only two order processors that have contact with OLM and these two were
interviewed. The main differences between them are that one is more experienced and handles
both short and long products while the other one only handles short products. According to the
order processor responsible for short and long products there are no control documents to
follow, which means that no standardization is followed. The order processors work procedure
differs a bit from each other’s and some steps and decisions are taken independently, especially
the one handling both short and long products. The following text and flow charts only concerns
new and test products, since the standardized products are handled automatically through the
EDI system.
36
rengiseD |
Chalmers University of Technology | Empirical Data
For the order processor that handles short and long products the work is initiated when
updating the list of new articles (see Figure 4-‐9). This list shows all the new products that R&D
has developed and that needs to be produced. The new products are thereafter put into a T-‐line
(manufacturing line), which shows what type of machines the product will go through. This
information is partly the base for the order processors to determine if the product will be
produced in-‐house or at OLM.
After the article update a CAD-‐drawing is received from R&D for approval and if everything is ok
an e-‐mail with approval is sent to the responsible designer. The designer hence releases the
drawing by uploading it to team center, which is a database for all the new drawings. The order
processor thereby downloads the released drawing from team center and sends it to OLM if the
decision to produce at OLM has been taken. A confirmation is received from OLM by mail if
everything is ok and then an operation list is created, including OLM in the operations list, by the
order processor and finished when a quote from the purchaser is received. When this is done an
approval is sent by e-‐mail to the planner, which has the authority to change the location of
manufacturing (in-‐house vs. OLM) depending on the capacity use in-‐house.
Other factors that are significant for the decision of where to produce are the quantity of orders
per month and if Sandvik Mining is capable of producing this product. The time spent on
communicating with OLM besides the e-‐mailing and confirmation is minimal and only occurs if
there are any corrections regarding orders. In the flow chart below the order processor´s work
procedure towards OLM is illustrated.
37 |
Chalmers University of Technology | Empirical Data
the order is received it is decided in collaboration with the production technique and planning
department where the product will be produced. Thereafter a CAD drawing is received from
R&D or downloaded from team center and sent to OLM to confirm that the drawing is ok. OLM
confirms the drawing through a phone call and then the order processor creates an operations
list, uploads it in the article register and confirms it for planning. When this is done an order
number is received from the planning department and the order processor creates an order.
Finally, a product cost calculation is made based on cost for similar products and assumptions.
The contact with OLM occurs when needed and not on a regular basis. What takes most time is
to create the order specification including the manufacturing line. Note that if Tempo 2 is
performed in-‐house instead of at OLM as initially decided, the goods receiving department might
be contacted. On the other hand when doubts about a drawing occurs the responsible designer
might be contacted. This order processors work procedure towards OLM is illustrated in the
following flow chart.
39 |
Chalmers University of Technology | Empirical Data
4.4.3 Production technicians
There are several production technicians involved with OLM but only two of them have been
interviewed. The production technician´s work procedure towards OLM is supposed to be equal,
however the flow charts differs from each other’s. There are no control documents for the
decision of what is going to be done at Sandvik Mining or OLM, instead the order processors´ and
production technicians´ experience and knowledge are used for this decision. Since the
production technician only have contact with OLM regarding new and test products it might be
difficult to have control documents describing each step since the procedure might differ from
product to product according to this production technician.
In Figure 4-‐11 one can see that the work procedure starts with OLM contacting a production
technician if any technical deviation is found. Primarily the order processor is contacted and if
he or she is available the order processor contacts the production technician if his or her skills
are needed. If possible the production technician gives a solution immediately, but sometimes an
investigation is needed and in that case the order processor is notified with a solution after the
technical investigation. If the production technician does not have the competence to solve the
problem it is forwarded to appropriate person e.g. the designer, and thereby the production
technician´s process ends.
In the case where the order processor not is available OLM contacts the production technician
directly for clarifications. If a solution is found it is given, but if not, an investigation is done or
appropriate person is contacted to get a solution. As an alternative OLM is forwarded to
appropriate person directly if it is found to be easier in that way.
41 |
Chalmers University of Technology | Empirical Data
Production technician
Sandvik Mining OLM
1. Order processor contacts
production technician if his or
Start her skills are needed.
2. If production technician do
not have the competence to
OLM contacts
solve the problem, it is
Sandvik AB if
forwarded to another person
any technical
and the production technician´s
deviations are
process ends.
found
OP
available? 3. Problem is solved and the
Yes No order processor is notified. if
no solution is found
immediately, appropriate
person is contacted and a
If OP is not solution is given to OLM. In
available PT is some cases the problem is
contacted for forwarded to appropriate
clarification person and department
directly.
OP contacts
production
technician (1)
Competence
to solve
problem? (2)
No Yes
Solve and
notify OP after
investigation
(3)
End
Figure 4-‐11: Production technician process.
The second production technician has the same role towards OLM but for the short products.
However he described his role as a support function. As with the former production technician
the process starts with OLM contacting the production technician, which in turn gives OLM
support. It could be questions regarding a drawing or minor errors that the production
technician can solve, otherwise the responsible designer is contacted to correct the drawing. If
no solution can be given OLM is directed to appropriate department and thereby the process for
the production technician ends. This production technician´s work procedure is illustrated in
Figure 4-‐12.
The contact between OLM and the production technicians occurs a few times per month and
mostly by phone. There are existing work procedures regarding how to work when new
products are being brought up, however some of them are still in progress.
42
naicinhcet
noitcudorP |
Chalmers University of Technology | Empirical Data
Production technician
Sandvik Mining OLM
1. Support is given and if
necessary OLM is referred to
Start appropriate department e.g.
the purchasing or R&D
department.
Contacts
Gives OLM
Sandvik Mining
support (1)
for support
End
Figure 4-‐12: Production technician process 2.
4.4.4 Planner
Concerning OLM there is only one planner and this person receives order from the order
processors. These orders are already planned either to be produced at OLM or in-‐house.
However the planner has the right to change an order from being produced at OLM to be
produced at Sandvik Mining if capacity is available. Regardless of this decision an order is
created in the ERP-‐system at Sandvik Mining. If the order is going to OLM an e-‐mail is sent and if
no answer is received it means that the order is confirmed. If the order is to be produced at
Sandvik Mining the process ends after the creation of the order in the ERP-‐system.
A lot of time is spent on changing, mixing and prioritizing orders, and sometimes the due dates
needs to be changed. The planner always tries to fill up the capacity use at Sandvik Mining and
since there is no existing work procedure for how to handle OLM´s products this is the only
guideline that is followed. The planner is also the only one that formally has the executive
authority to contact OLM and prioritize orders. When an order status needs to be confirmed the
planner might contact the goods receiver. Figure 4-‐13 illustrates the planner´s work procedure.
43
naicinhcet
noitcudorP |
Chalmers University of Technology | Empirical Data
Figure 4-‐13: Planner process.
4.4.5 Goods receiving
The goods receiving process is the procedure that occurs when material is moved between
Sandvik Mining and OLM. In Figure 4-‐14 below the process of goods receiving is illustrated. The
first step in this process is the receiving of material from the raw material suppliers, Ovako and
Tibnor. The materials received from these suppliers are ordered when the planner starts a
work-‐order towards OLM. The ordering is done automatically via the ERP-‐system when the
planner starts the work-‐order, however this order is not visible for OLM. After receiving the
material the next step is to register the material in the ERP-‐system and by this an order is
generated and sent to OLM automatically.
After registering the following step is to send the material to OLM for processing (Tempo 1)
which is scheduled to take 15 days for long product and 10 days for short products. After Tempo
1 the material is sent back to Sandvik Mining. When arriving to Sandvik Mining the material is
registered again, order cards are printed and material is sorted out according to destination. The
order cards are printed out in purpose of serving as information for the following processes.
The next step in the process is heat-‐treatment of the material. After heat-‐treatment, material
that is going to OLM for Tempo 2 returns to the good receiving and before being sent to OLM
there is a process of administration. Order cards that were printed are folded and order sheet is
printed. The order sheet is filled in manually with article number, quantity, price, drawing
number and desired due date. This physical order sheet and a manufacturing list, containing
information of what is needed to be processed on the material, is sent together with the material
44 |
Chalmers University of Technology | Empirical Data
to OLM for Tempo 2. After being processed at OLM in Tempo 2 the material is sent back to
Sandvik Mining where it is received and registered at the goods receiving.
There are no existing documents regarding work description for the goods receiver. All the
routines are based on the good receiver´s experience and nothing is written down.
Goods receiving
Sandvik Mining OLM
1. Material from OLM received and
registered. Order cards are printed
Start and the material are sorted
according to their destination.
2. Material which is supposed to go
Material from to OLM for the second tempo is
supplier heat treated.
received
3. Order cards foldered and order
sheet printed. Article number,
quantity, price, drawing number
Material
and desired due date is filled in on
registered in
order sheet. A manufacturing list is
ERP and order
also printed out and sent together
sent to OLM by
with the order sheet and material to
EDI
Tempo 1 OLM.
(4)
4. Material is processed (tempo 1)
Material
handling 5. Material is processed (tempo 2)
process (1)
OLM material
sent to heat
treatment (2)
Order
administration
(3)
Tempo 2
(5)
Material from
OLM received
and registered
End
Figure 4-‐14: Goods receiver process.
4.4.6 Purchaser
The purchaser works with development of suppliers, price negotiations, legal agreements and
other important issues with suppliers. However with OLM there are no legal agreements. When
quality or delivery precision is fading the purchaser may also get involved. The work procedure
45
gniviecer
sdooG |
Chalmers University of Technology | Empirical Data
for the purchaser towards OLM is vague, therefore no flow chart has been created for this role.
The purchaser and planner sit within the same department and works closely with each other
when it comes to OLM. The purchaser also has contact with OLM and the quality manager,
though very seldom.
In some cases OLM contacts the purchaser for instance when orders are late or when it comes to
questions regarding price or lead times. As for the most of the other roles that have been
interviewed regarding OLM there are no control documents for the purchaser to be followed.
4.5 Quality insurance process
In this sub-‐chapter the quality process is presented and further answers a part of the first
research objective under the third research question. The quality insurance process consists of
the following roles:
• Quality engineer
• Quality controller
• Measuring operators.
The quality department has contact with OLM and several sub-‐departments at Sandvik Mining
regarding OLM. Departments contacted are R&D, purchasing or goods reception department
depending on the type of issue. The contact regarding OLM is initiated when four issues occurs
which are the following:
• When Sandvik Mining finds deviation in quality at their site.
• When there is a problem concerning a drawing.
• When OLM finds deviation on the products at their site but do not know if it still is
acceptable.
• When different tools are in need of calibration.
Depending on type of issue different persons at Sandvik Mining are involved, these four issues
will be further presented below with text and flow charts under the headlines of respectively
responsible person.
4.5.1 Quality Engineer
As showed in Figure 4-‐6 the quality engineer is in contact with the quality controller, designers,
purchaser and planner regarding OLM. There are two situations that initiate the quality
engineer’s work towards OLM. The first situation is when OLM finds a problem with the drawing
that has been received about a new product. It starts with a detection of error on a drawing,
contact is then initiated with the quality engineer. There can be different kind of problems
regarding the drawing. There can be a situation where OLM already has produced parts and then
realized that there is an error. The other situation that might occur is that the error is identified
on the drawing before production.
Both issues have more or less the same solving-‐process, regardless of how serious the problem
is. As illustrated in Figure 4-‐15 below a quality deviation is detected and contact is initiated with
the quality engineer. If it shows to be a minor issue that can be solved instantly the quality
engineer notifies OLM about what to change.
If the issue is of major concern then an investigation is done at Sandvik Mining in order to locate
the reason for this problem. When a solution and counter-‐measure is found, OLM is contacted
and given the information about this in order to prevent it from happening again. In these
situations it might happen that the quality engineer is not the adequate person to identify a
solution for the problem. If so is the case, the problem is delegated to the person who has the
46 |
Chalmers University of Technology | Empirical Data
knowledge to solve it. Most often these persons are the designers since these are the ones
sketching the drawing and therefore know what the intentions from the beginning were.
Quality engineer: New products
Sandvik Mining OLM
1. If it is a minor problem
without the need of further
Start actions, OLM is told that it is
okey for this time but that it
has to be correct next time.
If further actions are needed
Quality
an investigation is made.
problem
Thereafter the whole process
detected?
ends. But if there is a major
problem a more thourough
investigation is done in order
Sandvik Mining to find a solution.
informed about
the problem 2. OLM contacted with the
through phone solution and counter
measures are given.
Investigation of
quality problem
(1)
Solution given
to OLM (2)
End
Figure 4-‐15: Quality engineer process when problem detected on new products at OLM.
The second situation that might occur that directly affect the quality engineer is when OLM
identifies deviation in quality on physical products after being processed but do not know if they
still are acceptable. When this occurs OLM contacts the quality engineer about the deviation and
sends the products to Sandvik Mining. When Sandvik Mining receives the products these are
controlled and the decision about what to do with the products is taken. If the decision is to
scrap the products, OLM is notified about the decision, and in that case OLM is not compensated.
The purchaser is thereafter notified about the situation so that statistics can be registered about
the supplier. The planner is also notified, who thereby takes the decision if any counter-‐
measures are needed is taken e.g. release a new order or prioritize existing orders. The process
is illustrated in Figure 4-‐16.
There is also the situation where the products can be fixed and when this occurs the products
are fixed and both OLM and the purchaser are notified about the change. If the deviation is too
much but the product is still acceptable a new price is negotiated.
According to the quality engineer these processes summarize most of the problems that occur,
but the quality manager also said that every specific problem is unique and that the problem
solving that the quality engineer applies in order to solve these issues is something that is based
47
reenigne
ytilauQ |
Chalmers University of Technology | Empirical Data
Quality controller
Sandvik Mining OLM
1. Whether there is a
deviaiton in quality or not it
Start is registered in the system
and thereby the process
ends. If there is a deviation
the quality engineer is
notified before ending the
Sampling process.
Yes No
Deviation in
quality
Register in
system(1)
Notify quality
engineer
End
Figure 4-‐17: Quality controller process.
During this process, the first thing that happens is to quality check a sample of the received
material. If the material passes the control the next step is to register this in an excel-‐file and no
further action is taken. If there is a deviation in quality the next step is to register this in the
excel-‐file and later notify the quality engineer about it and the quality controllers process ends.
Currently there are no control documents regarding the work for the quality controller, but
there are existing control documents regarding how often suppliers products are going to be
sampled during a specific period of time.
4.5.3 Measuring department
Sandvik Mining provides OLM with equipment for the quality control performed at OLM, and
when the expiration date on these tools has expired the tools have to be sent to Sandvik Mining
for calibration. In Figure 4-‐18 an illustration of this process is presented.
49
rellortnoc
ytilauQ |
Chalmers University of Technology | Empirical Data
Measuring operator
Sandvik Mining OLM
1. Tools received and
Start calibrated. Thereafter it is
registered in the system.
Expiration
date
Notification Notification
sent to OLM received
Calibration and Tools sent to
registration (1) Sandvik Mining
Tools sent to
OLM
End
Figure 4-‐18: Measuring operator process.
Sandvik Mining keeps track of the expirations dates and when the time comes Sandvik Mining
sends a notification physically together with the daily deliveries to OLM, informing about which
tools that have to be sent to Sandvik Mining for calibration. Further the tools are sent to Sandvik
Mining and after the calibration the tools are sent back to OLM.
Currently at the measuring unit there are no existing control documents regarding the practice
of the work that is performed at the measuring unit but according to the operators this is in
progress.
4.6 Summary of empirical data
Following is a brief summary of what have been presented in this chapter.
At first the material flow between Sandvik Mining and OLM was presented. As illustrated in
Figure 4-‐3 Sandvik Mining orders material from raw material suppliers (Ovako Forsbacka,
Ovako Hällefors and Tibnor), with a delivery time of 5 days. After receiving and registering the
material it is sent to OLM for Tempo 1, which has the lead time of 16 days. After Tempo 1 the
material is sent back to Sandvik Mining for heat treatment where 84.3% of the material will be
left at Sandvik Mining for further processing while the remaining 15.7 % of the material is sent
back to OLM for Tempo 2. The annual weight of material sent to OLM is approximately 5060
tons.
50
rotarepo
gnirusaeM |
Chalmers University of Technology | Empirical Data
The second part that was presented was the information flow. Firstly a chart (see Figure 4-‐5)
including all of the departments at Sandvik Mining that are involved with OLM was presented in
order to get an overview of how the communication looks. Later on a more thorough chart was
presented (see Figure 4-‐6) showing how the communication, on an individual level looks like
internally at Sandvik Mining and externally towards OLM.
Lastly, the roles that are involved with OLM were presented more thoroughly. Where their work
procedure is presented in form of block-‐diagrams in order to get an understanding of how the
work looks on an individual level. Also presented was if there are any control documents
regarding the work procedure.
The following table (Table 4-‐3) is a summary of which of the roles that have control documents
regarding their work procedure. As seen in the table it is showed that the only roles that are
involved with OLM that has some kind of control documents regarding their work procedure are
the order processor for short products and the production technicians to some extent. The
production technicians have control documents regarding some processes of their work. The
rest of the roles do not have any control documents meaning that all of the work performed by
these persons is inherent. Furthermore, no steering group or formal boundaries of authorities
exists, enabling everyone to contact everybody. This answers the second objective in the third
research question in the purpose chapter.
Table 4-‐3: Existing control documents for each role involved with OLM
Role Control documents
Designer None
Designer None
Order processor short & long products None
Order processor short products Existing
Production technician Existing to some degree
Production technician Existing to some degree
Planner None
Goods receiving None
Purchaser None
Quality engineer None
Quality controller Existing to some degree
Measuring operator None
51 |
Chalmers University of Technology | Analysis
5 Analysis
All the figures and flow charts that have been described to this point stands as a base for the
decision, concerning an elaboration of the areas where potential for improvement were
identified. Three main areas were found to have potential for improvement and are therefore
further analyzed in this chapter whereof one is presented as a business case. The upcoming data
and calculations in this chapter are based on internal data provided by Sandvik Mining.
The business case will show the possible cost savings in a contingent case where the material
that are supposed to go to OLM goes directly to OLM from the raw material suppliers, instead of
coming to Sandvik Mining for unloading/ loading and thereafter going to OLM a after day or two.
This contingent case would incur a few hours less work per month for the goods receiving staff
handling material, while it would incur a few minutes extra work for e.g. the order processors
and planner. This amount of time is marginal and has therefore it has not been considered in any
cost saving calculations.
The second improvement area that is analyzed concerns the information flow i.e. how should the
information flow be structured towards OLM, which steering groups should exist and how
should the information be transmitted.
The last area with potential for improvement that was analyzed is about the processes and how
these can be improved e.g. by standardization and by having control documents. The chosen
elaboration areas were decided in collaboration with the supervisor and purchasing manager at
Sandvik Mining.
5.1 Material flow analysis
The business case covers the main cost savings in the contingent case where OLM buys the
material. However the whole business case is based on forecasts and incoming material to OLM
during February 2013 and April 2012, however the two months differed a bit. February was
chosen since it was the most recent month at that specific time and April was chosen since it is
not affected by any vacations or fluctuations in the demand due to season.
OLM is currently seen and treated in many ways as a part of Sandvik Mining even though it is an
external supplier. The information sharing between OLM and Sandvik is quite poor and OLM
receives an order a few days before the material is sent to OLM. OLM receives an order with a
maximum of one week in advance from Sandvik Mining. This might be a contributing factor for
why OLM has low delivery precision and long lead times since they only can plan one week in
advance.
As mentioned earlier new and test products correspond to 2% of all orders sent to OLM and
these products have the highest uncertainty and longest production lead time. The risk with this
uncertainty is that the new and test products might interrupt the production at OLM causing the
low delivery precision and extending lead time for standard products. Another aspect that might
affect the delivery precision are that no forecasts are shared with OLM resulting in planning
issues. In addition, due to that some suppliers deliver the material too early and that there is a
lack of space, Sandvik Mining sends the material as soon as possible. This also means that the
material arrives too early at OLM, hindering them to plan and prioritize their production in an
efficient way, which also could be a contributing factor to the poor delivery precision and
inaccurate lead times.
According to Lyson & Farrington (2012) collaboration and information sharing is vital in order
to achieve a well-‐functioning supply chain. Therefore, it is important that these two things work
properly. Segerstedt (2009) means that SCM strives to reduce costs and increasing value
creation through improving the whole supply chain. With these two statements in mind and
with the investigation of the material flow it has been identified that it could exist a great
54 |
Chalmers University of Technology | Analysis
potential for improvement by treating OLM as a ordinary supplier and not as a part of Sandvik
Mining as it currently is.
The main reasons for why OLM should become a more ordinary supplier is in order to create a
more straight forward material flow, making a clear distinction that OLM is a supplier and not a
part of Sandvik, triggering OLM to develop their own organization and that Sandvik Mining will
not be the owner of the material until it is received after Tempo 1. In addition, as mentioned in
the empirical data there is not enough space to store the material, which also is a reason for why
OLM should purchase and receive the material directly from the raw material suppliers.
Several wastes in the current material flow for material going to OLM were identified such as
extra waiting time and excess inventory. Liker & Meier (2006) argues that work in progress, raw
material that causes obsolescence, transportation and storage cost, longer lead times or delays
are caused by excess inventory, and several of these could be identified in the current situation
with OLM.
The material coming from the suppliers that are supposed to go to OLM are first sent to Sandvik
Mining where it is unloaded, sorted and stacked and after a day or two sent to OLM for Tempo 1.
If the material instead would be sent to OLM directly from the raw material suppliers the
unloading, sorting and stacking before Tempo 1 would be eliminated and concurrently the lead
time would be reduced for Sandvik Mining with approximately 16 days. This since OLM would
have to buy the material, owning it until it has been delivered to Sandvik Mining after Tempo 1.
At the same time the tied up capital would be reduced with approximately 16 days since Sandvik
Mining would not own the material until after Tempo 1. In addition, a more ordinary supply
chain would be created with OLM, which as mentioned earlier could have a great potential for
improvement. Furthermore, if OLM would have access to more information such as forecasts it
could result in that OLM starts planning their production in a longer perspective enabling the
lead times to be reduced and more accurate. This would reduce the time between customer
order and delivery to customer, which together with eliminating waste is the basic idea with
lean production.
5.1.1 Material flow – business case
The business case will show the potential cost savings that exists if the material would go to
OLM directly instead of coming to Sandvik Mining before Tempo 1. A great amount of data were
gathered and analyzed in order to be able to come up with all the necessary data that were
needed in order to calculate and estimate a possible cost saving in the contingent case where
OLM buys their own material. Figure 5-‐1 below illustrates the material flow in the contingent
case. Note that only 15.7% of the material would go to OLM for Tempo 2.
OLM OLM
Tempo
2
Supplier Sandvik Mining Sandvik Mining
Figure 5-‐1: Illustration of the material flow in the contingent case where OLM buys the material
55 |
Chalmers University of Technology | Analysis
There were two main costs identified in the current situation that could be reduced, which are
the transportation and holding cost. The holding cost is in turn based on tied up capital, which
also was identified as a potential for cost savings.
5.1.1.1 Holding cost savings and released tied up capital
All the steel prices (in SEK) were calculated based on price lists for different items from the
three raw material suppliers. The price list that was used was provided by the company but is
not presented in this report.
Based on the material cost per kilogram and the forecast for March, April and May (see Table
5-‐1) the material cost for the material going to OLM has been calculated and is illustrated in
Table 5-‐2. The last row in these tables shows what the demand respectively material cost for a
whole year ought to have been if the monthly demand would be the same through the whole
year.
Table 5-‐1: Forecast in kilograms of material going to OLM
Month Mar-‐13 Apr-‐13 May-‐13
Forecast 129 829 132 343 135 546
Sum/ year 1 557 953 1 588 117 1 626 548
Table 5-‐2: Total material cost based on forecast and material price
Month Mar-‐2013 Apr-‐2013 May-‐2013 Average
Cost 2 555 963 2 607 476 2 668 975 2 610 804
Sum/ year 30 671 555 31 289 709 32 027 696 31 329 653
The average material cost for a year is calculated to 31 329 653 SEK, which means that since the
lead time at OLM is calculated to be 16 days on average, Sandvik Mining has 16 days of tied up
capital. The value of material for these 16 days is calculated to 1 373 354 SEK (see calculation
below).
31 329 653 SEK
∗16 = 1 373 354 𝑆𝐸𝐾
365
This is the value of the material that Sandvik Mining owns from the moment of receiving the
material from the raw material suppliers before Tempo 1 until the material is received again
after Tempo 1. Since material comes every day from the suppliers, the level of material always
stays on top, meaning that the average value of material during a year is the top value.
Noteworthy is that the tied up capital will only be released once. Since the holding cost interest
at Sandvik Mining is 10%, the yearly holding cost saving would be 137 335 SEK, see calculation
below
1 373 354∗0.1 = 137 335
5.1.1.2 Transportation cost savings
The transportation cost was also identified as a potential for cost saving. As mentioned before, in
the current situation Sandvik Mining hires one truck from a 3PL company on full time, which
among other things picks up material from Forsbacka twice a day and delivers and picks up
material at OLM for Tempo 1 and 2 once a day. This truck will still have the same amount of trips
and deliver the material from Ovako Forsbacka to OLM through Sandvik Mining. In this
contingent case OLM will be the owner of the material and therefore gives Sandvik Mining the
opportunity to negotiate a lower price, since Sandvik Mining already pays for the transportation.
56 |
Chalmers University of Technology | Analysis
Another aspect is that material coming from Ovako Hällefors and Tibnor will not go through
Sandvik Mining anymore. With less material being sent to OLM by the truck hired from Sandvik
Mining, gives the opportunity to revise the usage of this truck and see if there is a possibility for
cost savings. However, this cost saving has not been included in the business case due to the
complexity to investigate.
Nevertheless, there are two other transportation costs, one from Tibnor and one from Hällefors
that that has been taken into consideration since the material coming from these suppliers are
delivered in other trucks. These transportation costs are calculated below.
Internal data shows that Tibnor, which only produces one out of five steel types sent to OLM,
charges 4.79 SEK/ kilogram in transportation costs and since 88% (51 696kg/ year) of the
material weight sent from Tibnor is supposed to go to OLM, it means that there is a great
potential for cost reduction. The calculations below show how much that can be saved in
transportation costs from Tibnor in the contingent case.
51 696∗4.79 = 247 624 𝑆𝐸𝐾 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟
Another aspect that needs to be considered with the new material flow and Tibnor is that the
transportation cost is based on weight and a minimum weight is required to obtain the current
price. Since 88% of the material coming from Tibnor goes to OLM it is of essence to consider the
new transportation cost for the remaining 12%, i.e. approximately 500 kg/ month that Sandvik
Mining still will receive.
The annual weight of material coming from Hällefors is 942 588 kg. Giving that the
transportation cost from Hällefors is 0.20 SEK/ kg it would give a total transportation cost
saving of 188 518 SEK/ year, see calculation below.
942 588 ∗0.2 = 188 518 𝑆𝐸𝐾
The total transportation cost saving for transports from Tibnor and Hällefors is:
247 624+188 518 = 436 142 𝑆𝐸𝐾 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟
The annual cost savings from both the holding cost and transportation cost gives the sum of:
436 142+137 335 = 573 477 𝑆𝐸𝐾 𝑝𝑒𝑟 𝑦𝑒𝑎𝑟
Finally, it is of essence to consider OLM’s new cost, which has not been calculated since there are
many factors to take into consideration. Furthermore, it was not within the scope of the purpose.
5.1.2 Material flow summary
Due to that the calculations are based on forecast the results from the calculations have been
rounded to an annual saving of 570 000 SEK and released tied up capital to 1 370 000 SEK
Besides the cost savings it is important to consider that if OLM buys the material it means that
Sandvik Mining will be charged a higher price. This due to that OLM not only will charge Sandvik
Mining for the processing cost but also for the material cost and all extra costs such as
transportation and holding cost that supervene if OLM buys the material. Initially it will not
create any value for OLM, however it should trigger OLM to develop their planning which is a
step towards reducing the lead times and increasing the delivery precision, which is a benefit for
both companies and the purpose of a supply chain according to Christopher (2011).
The main advantages with the contingent case are that Sandvik Mining will release tied up
capital due to the reduced lead time as well as reduce the holding and transportation costs.
Other advantages are that OLM will be considered as an external part by becoming a more
ordinary supplier, which purchases the material and by this creating a more straightforward
57 |
Chalmers University of Technology | Analysis
supply chain. Furthermore, it should also trigger OLM to develop their organization since OLM
no longer will be treated as a part of Sandvik Mining. Additionally the goods receiving
department at Sandvik Mining will not handle the material before Tempo 1, which for the
moment is done and can be seen as a waste. If OLM would receive forecasts it would enable them
to plan their production in a longer perspective and thereby possibly reduce the lead time for
the whole supply chain as well as gaining economies of scale. Lastly, the structure towards OLM
will be as a more ordinary supply chain enabling Sandvik Mining to work with OLM as a
ordinary supplier, improving the delivery precision and continuing eliminating waste from an
improved position.
5.2 Information flow analysis
5.2.1 Supply chain information flow analysis
As mentioned earlier Sandvik Mining should consider making OLM a more ordinary supplier. In
order to have a well-‐functioning supply chain the information flow and technology between OLM
and Sandvik Mining should become better, since technology is one of the most important
enablers for a well-‐functioning supply chain (Lyson & Farrington, 2012). According to
Vanpoucke et al (2009) there is a relation between the amount of information shared and the
overall performance of the supply chain, supply chains with less information sharing perform
poorer in comparison to supply chains that use more information sharing.
5.2.1.1 Internal data analysis
As mentioned earlier Sandvik Mining uses fixed lead times for their standard products when
being sent to OLM for processing even though these lead times are not updated to the actual lead
time. Long products have a lead time of 15 days, short products 10 days and product sent for
Tempo 2 has 10 days as well. By letting the ERP-‐system to set the end dates to specific days (not
only on Fridays), and set the lead times more accurately i.e. set specific lead time for every
product type, the delivery precision should be improved. Furthermore, the theoretical lead time
would also be decreased since the end dates not will be extended to the next Friday in that
specific week. This means that the lead time for the next operation can be planned to start up to
four days earlier.
Therefore, the recommendation for Sandvik Mining is to revise the existing information data,
such as the lead times, making the lead times more accurate and adjusted to specific product
types. By updating this information it generates the possibility to improve the supply chain since
more accurate data will be used.
5.2.1.2 Inter-‐company communication analysis
The communication between OLM and Sandvik Mining occurs through phone, e-‐mail, ERP-‐
system and physical meetings. Lyson & Farrington (2012) state that one of the most important
enablers of a well-‐performing supply chain is achieved by having compatible IT-‐systems. This is
not the case with OLM and Sandvik Mining in the current situation. The ERP connection between
OLM and Sandvik Mining consists today of that an automatic order is generated and sent to OLM
when Sandvik Mining receives the raw material. The order is then printed out at the site at OLM
and manually registered into the ERP-‐system at OLM.
Also notable is that the forecast that OLM receives from Sandvik Mining is as a maximum one
week in advance, making it difficult for OLM to plan their production. Improving the
collaboration between companies by using tools such as CPFR will give the possibility to
automatize different processes between the companies such as sharing information about
planning, forecasting and replenishment. Fliedner (2003) states that the supply chain benefits of
using CPFR are improved forecast accuracy, lower system expenses and reduction of stocking
points (making a more direct material flow). By reducing the number of stock points the amount
of inventory in the supply chain will decrease which in turn will shorten the total lead time of
the supply chain (Srinivasan, 2004). A compatible IT-‐system will also automatize the ordering
58 |
Chalmers University of Technology | Analysis
and billing, without any manual labor. In addition, according to Vanpoucke et al (2009) a more
sophisticated information sharing between the companies could improve the supply chain
coordination and decrease the supply chain cost, which is desired in this case
5.2.2 Internal structure analysis
As presented in the chapter “Empirical data” it is identified that there are different divisions
with several employees at Sandvik Mining that are involved with OLM, matter of fact there are
approximately 50 employees that are in contact with OLM.
The organizational structure that is present at Sandvik Mining, though it is not formal, can be
identified as Galbraith et al (2001) claims a product form, where the company is divided
according to products. This since it was identified that Sandvik Mining had multiple contact
points with OLM, divisions working independently, and that there were a poor internal
collaboration between the divisions, which might hinder the possibility to achieve benefits such
as economies of scale (Shani, Chandler, Coget, & Lau, 2009). Economies of scale could be
achieved by placing orders with similar products next to each other’s for instance. This is only
possible if there is collaboration and mutual planning at Sandvik Mining, which currently is poor.
Due to that Sandvik Mining was identified to have a product organizational structure with many
employees contacting OLM results in an issue regarding prioritization of orders. Furthermore,
there are a lot of different employees with different information resulting in difficulties for OLM
to cope with all of this information. As Lyson & Farrington (2012) states the most important
enabler of a well-‐functioning supply chain is the organizational infrastructure. Furthermore, it is
important to have a structure that facilitates to follow the strategy and also to have an
organizational structure of business units and functional areas that suits the company in order
to enable a smooth supply chain. Hence the need to build an organizational structure in order to
create a more clear and straight forward information sharing between Sandvik Mining and OLM.
In some cases a functional organization structure is more efficient than organizations with a
product organization structure Shani et al (2009). Having a more functional structure interface
towards OLM would generate a more straightforward information sharing and a clearer
hierarchy of which persons that has the authority when it comes to prioritizing and decision
making.
5.2.2.1 Standard products
The process flow charts in the empirical data chapter showed that most of the communication
occurs when new and test products are being developed or produced. The communication
regarding standard products occurs automatically without any human intervention. Therefore,
two different information flow structures have been developed depending on the type of
product that will be produced. Since all communication occurs automatically when it comes to
standard products there is no need for human intervention between some departments, hence
the possibility to only allow a few individuals in the new information flow structure to have
contact with OLM for standard products. Figure 5-‐2 below illustrates an alternative structure for
the communication structure towards OLM between departments for the standard products.
59 |
Chalmers University of Technology | Analysis
While the current communication structure towards OLM is vague, this alternative one is a
functional structure which according to Galbraith et al (2001) has the advantages of knowledge
sharing, specialization and standardization. One of the disadvantages might be that there is a
lack of cross-‐functional processes but since the standard products rarely causes any problems
there is no risk for that. If necessary the different roles at Sandvik Mining should have the
possibility to contact each other even though this not might happen too often. Nevertheless, the
knowledge and information sharing towards OLM is vital in this case which is achieved with a
functional structure and which also are some of the advantages of having a functional structure
according to Galbraith et al (2001). This structure should lie as a base for all lego-‐suppliers in
order to create an identical and standardized interface towards all lego-‐supplier.
5.2.2.2 New and test products
There are more departments involved with OLM when developing new and test products.
Therefore, it is of great importance to keep the communication between OLM and these
departments because of the importance of OLM´s involvement in the development process. This
means that everybody at Sandvik Mining should be allowed to contact whoever is necessary as
long as it is for development issues.
As mentioned earlier the current communication structure (see Figure 4-‐6) can be seen as a
product structure which for the new and test products might be the most appropriate one. Since
two different structures are proposed depending on product type the main structure for Sandvik
Mining will be a hybrid structure. This structure allows the employees to follow the functional
structure for standard products and the product structure for new and test products. However
regardless of structure a clear division of authority towards OLM is also recommended in order
to create an efficient communication flow. The steering group created for the standard products
would be the same for the new and test products.
5.2.2.3 Steering group
A steering group is proposed consisting of the planner and order processors which ought to be
the ones managing the relationship with OLM for all products. This since they already are the
ones deciding what to produce at OLM in the current situation and also because they are the
ones with the necessary knowledge and data to follow the three priority rules, which still will be
followed. For instance the order processors are the one with the knowledge of what can and
should be produced at OLM while the planner is the one who has the overall view of the capacity.
However the planner should be the one with the executive authority. Since the standard
products stands for approximately 98% of all the orders sent to OLM it is of essence that a clear
and organized structure is emphasized and followed for these at Sandvik Mining.
Furthermore, the steering group has the purpose to create a united Sandvik Mining front
towards OLM. The steering group should when necessary, during meetings discuss what should
and can be produced at OLM during a specific period. Sandvik Mining should use existing
information at to create forecast and solve issues that occurs with OLM. Since the
administrational handling of standard product is automatized there is not a high human
interference in that flow. Therefore, the meetings should occur when there is a need of new
forecasts, priority issues and decisions regarding where to produce (Sandvik Mining or OLM) or
other kind of deviations that needs to be solved.
5.2.2.4 Goods receiver, quality manager and purchaser
The goods receiver and quality manager should also be allowed to have direct contact with OLM
to a limited extent, since some issues only regards these roles e.g. the calibration of tools and
quality problems detected at either OLM or Sandvik Mining. This direct contact enables these
persons to handle an issue within their specialized area such as a quality problem direct with
OLM, mitigating the risk of dropping information due to a longer information chain. The
specialization is one of the advantages with a functional structure and is therefore another
61 |
Chalmers University of Technology | Analysis
reason for having a functional structure. The roles at the bottom of the structure may not be
involved with OLM for standard products but should not be excluded.
Lastly, the purchaser should also have contact with OLM regarding negotiation about pricing and
legal agreements but also perform follow-‐ups when there are deviations in delivery precision
and quality etc. The purchaser should only discuss the issues mentioned above and not the
products, priority or product status per se. The purchaser has therefore been placed at the same
level as the planner but with focus on other kind of questions.
5.2.2.5 Designers, production technicians and quality manager
The communication with OLM from designers, production technicians and quality manager
should only consider development and quality issues i.e. each role should have clear boundaries
and responsibilities. The goods receiving department will still have the paper sheet transferred
to OLM for Tempo 2, even though this in the future could be improved with a more sophisticated
ERP-‐system, which already is in progress. Furthermore, the steering group should consist of the
same role i.e. the planner and order processor, regardless of structure. As mentioned earlier the
planner should be the only one with the authority to prioritize orders since this person is the
most appropriate one for this purpose regardless of structures.
5.2.3 Information flow summary
As stated earlier the information sharing between OLM and Sandvik Mining is poor. In order to
enable the information sharing with OLM, Sandvik Mining should consider developing their ERP-‐
system collaboration with OLM. Today the communication occurs through phone, e-‐mail, mail or
physical meeting and a lot of these processes can be excluded and automatized by a more
sophisticated ERP-‐system collaboration. This collaboration should generate benefits for the
supply chain as whole in terms of shorter lead times and more accurate delivery precision.
The data that Sandvik Mining is using is too inaccurate, the data should be revised and used on a
product type level. In order to increase the accuracy of the information and reduce the
theoretical lead time, the end-‐dates, which today are set to Fridays, should also be revised and
changed so that end-‐dates can be set to specific days.
Finally, the organizational structure should be revised and looked upon. Sandvik Mining has
been identified to have a product group organization, and one of the cons of this organization is
that there are too many contact interfaces present in the relationship between OLM and Sandvik
Mining for standard products. These interfaces can create priority issues at OLM resulting in low
delivery precision. Therefore, a new organizational structure has been suggested for standard
products, a more functional one, where there is a clearer division of the organization and
authority. This structure should be the formal one for standard products while the structure for
new and test products should be the same as the current one. Both structures should be
emphasized as the formal ones depending on product type i.e. a functional structure for
standard products and a product structure for new and test products i.e. a hybrid structure.
Furthermore, the new structures should lie as a base and be possible to be applied for all lego
suppliers in order to create an identical and standardized interface towards all lego suppliers.
The main differences for the different roles due to the change in the communication structure
towards OLM would be that a steering group consisting of the order processors and planner
(which takes the final decisions) is created.
62 |
Chalmers University of Technology | Analysis
5.3 Processes analysis
As Olhager (2000) states the analysis of the processes can vary in level of detail and information
depending on the purpose of the charts. In this case the purpose was to map the interfaces in the
information flow in the current situation. Therefore, detailed mappings of individual processes
were performed to identify all the interfaces towards OLM. The process activities were identified
and categorized and at the same time the whole process for each role regarding OLM was
documented which according to Olhager (2000) are the fundamental steps in a process flow
analysis.
It was identified that the amount of roles involved with OLM varies depending on type of
product that is referred. For the standard products there are almost no contact between
individuals since most of the communication occurs automatically through the ERP-‐system
while there are a lot of communication through e-‐mail, phone and physical meetings between
individuals when it comes to new and test products that are being developed. Irrespective of
product type there are no standardized ways of working or any control documents (except for a
few ones, see Table 4-‐3) towards OLM i.e. almost everything is based on know-‐how and
experience. According to Liker & Meier (2006) standardization is needed in order to be able to
compare possible improvements but if there are no standardizations there are no reference
points to compare with.
Also having no standardization makes Sandvik Mining vulnerable since almost nothing is written
down. For instance, if a new person is employed it is impossible for that person to follow any
guidelines for their work. In addition, the analysis showed that some individuals e.g. the
designers act differently in the same situation. This might be due to that there are no
standardized ways of working. Hence why the designers do what for the moment is most
comfortable, even though this might not be the most adequate. For instance, one designer
contacts OLM directly for order status while another designer contacts the planner for the same
type of question. This order status contact with OLM might be one of the reasons that cause a
priority issue for OLM.
Even though the new and test products only stand for approximately 2% of all the orders, most
of the contact occurs regarding these products, hence the importance of having control
documents and standardized processes. But Liker & Meier (2006) state that a main prerequisite
for standardization is that the tasks are repeatable. This could be an issue since every contact
and procedure during the development process might differ from each other even though the
contacts are repeatable.
5.4 Action plan
As mentioned earlier in the analysis, several improvements have been suggested for a better
relationship such as formalizing the organizational structure, standardize the work tasks and
making OLM a more ordinary supplier by changing the material flow. In Figure 5-‐4, an action
plan is presented which shows how the work should progress at Sandvik Mining in order to,
improve the accuracy of the lead times, reduce the lead times and improve the relationship with
OLM.
Firstly, it is recommended to standardize the work tasks, which also should be the base for the
creation of control documents and will be used as guidelines for every role. This process should
approximately take 6 months and be created by the different roles involved with OLM. Since
some roles involve several persons e.g. the designers and order processors, it is important that
these persons agree upon a common standard that all persons with this role follows. This
process ought to be controlled by the manager of each role to secure that it has been done.
Simultaneously the organizational structure should be formalized in order to create a structure
with clear division of authority at Sandvik Mining. This process does not have a formal end date,
it is a change that should be made and maintained in the future as well as emphasized for every
63 |
Chalmers University of Technology | Analysis
role involved with OLM. Therefore, it is important to involve the top managers consisting of the
production unit manager, purchasing & logistics manager, production manager, R&D manager
and sourcing manager and that this group reaches consensus and decides that the new structure
is the currently best one and thereby should be followed. The top managers should thereby
work as a supportive function towards the involved roles and secure that the new structure is
followed. It is important to achieve the top managers´ support since it otherwise will risk
phasing out and return to how it was done before. This is a top-‐down management way of
performing a change i.e. that the power to execute the change lies on the top management, which
pushes the change through the organization (Tsai & Beverton, 2007). Furthermore, OLM has to
be informed that the new structure is the formal one and that the planner is the only one with
the executive authority.
It is also recommended to initiate the revision of the data about the lead times, which currently
is recognized as inaccurate. The process of updating the lead times is seen as the beginning of
implementing a new material flow and is considered to take approximately 2 months. The next
process that follows is to negotiate new prices with OLM, which should take approximately 3
months. The purchaser in collaboration with OLM should perform the update and agreement of
new lead times and prices. Finally, the new material flow should be implemented, which is
approximated to take 6 months. The implementation should be held and chaired by the
purchaser.
Figure 5-‐4: Action plan
During this time, while these steps are executed there will be a development of the ERP-‐system.
The planning of a new ERP-‐system at Sandvik Mining has already been initiated and is a process
that involves Sandvik Mining as a whole and not only the relationship with OLM. This is a
process that is ongoing throughout the whole supplier development between Sandvik Mining
and OLM. Still it is of the essence to implement an ERP-‐system that will enable a more efficient
collaboration between Sandvik Mining and OLM. Moreover, the new ERP-‐system ought to be
able to be used equally with all suppliers but at the same time differently depending on type of
relationship.
64 |
Chalmers University of Technology | Discussion
6 Discussion
6.1 Is the new material flow the primary solution to the actual problem?
Can the real problem lie in the information sharing between Sandvik Mining and OLM? Our focus
has primarily been on the material flow between the companies. But, what if the real solution
lies in the information sharing between the companies.
Are the new and test products causing the low delivery precision? As we mention earlier
approximately 2% of all orders regards new and test products, and these can delay standard
products up to 4 days. The 16 days of lead time at Tempo 1 has most likely been affected
negatively by these 2%, since these products use to have a longer lead time than standard
products. Our opinion is that Sandvik Mining might have too little information about OLM’s
production and capacity making it harder for Sandvik Mining to know what and how much to
put on lego and which problems it causes OLM. According to Lyson & Farrington (2012),
information sharing is vital for a well-‐functioning supply chain, therefore it is important that the
information sharing between these two companies work properly in our opinion.
How does the prioritization affect the delivery precision? Sandvik Mining seems to have
problems planning the prioritization at OLM between standard and new and test products, and
that could be a reason why OLM has a low delivery precision. The problem with capacity usage
when producing standard and development products in the same production seems to be a
common problem for manufacturing companies with R&D departments, Macintosh (2007).We
are of the opinion that Sandvik Mining does not know how much their prioritization affects the
production at OLM. We think that prioritizations made by Sandvik Mining should not affect the
delivery precision i.e. if Sandvik Mining prioritizes a product causing a delay for other products,
it is important to emphasize that this is due to Sandvik Mining themselves and not due to OLM.
Can capacity reservation be a part of the solution? One solution to this might be to have
reserved capacity for these type of products which the R&D department can dispose as they
want. However the steering group should be notified when new and test products are being
developed and needs to be produced at OLM. This would enable the planner to plan the
production and prioritize orders more efficiently.
Maybe not all suppliers should be treated equally! It is also important to consider if OLM´s
delivery precision ought to be as high as other supplier´s which not produces complex products
as OLM does. It maybe should be allowed that OLM has a delivery precision with a wider range
of acceptance due to the complexity and quality of the products and the high service level from
OLM.
6.2 Will the new material flow be beneficial for all parties?
The purpose with SCM is that all parties involved gain benefits i.e. that a win-‐win situation is
created for all parties involved through collaborating. Furthermore, SCM is also about focusing
on one´s core competence and capabilities. Hence the reason to ask oneself, what it is that
differentiates one’s company from the competitors´. (Christopher, 2001)
In the contingent case, that is presented, where OLM buys the material instead of Sandvik Mining
the question that we ask ourselves is if this is a win-‐win situation for both companies. The risk is
that only one company gains benefit, or that both of the companies lose when going through this
change. It is easy to only go after what is measurable and miss other important things that are
more complex to put a monetary value on e.g. the R&D collaboration with OLM.
What if only Sandvik Mining benefits from the new material? It is important to remember
that OLM´s core competence is to produce a variety of complex products and small batches with
an almost excellent quality and service level. By initiating this change it forces OLM to start with
three new functions, which are purchasing, transportation and storing. A win-‐lose situation that
66 |
Chalmers University of Technology | Discussion
might occur in this change according to us is that OLM maintains the same quality and capability
to produce various products but the initiation of three new functions strikes harsh on the
company financially, while Sandvik Mining do not pay more. We think that it is important to
consider if it is necessary to treat OLM in the same way as an ordinary supplier; what if the
current material flow is the best one? OLM is maybe not a good performer in purchasing
material, storing material and planning transportations. OLM should maybe only focus on their
core competence which is to produce products that Sandvik Mining cannot, helping Sandvik
Mining when they are in need when it comes to capacity and support Sandvik Mining’s R&D
department.
What if only OLM benefits from the new material flow? A lose–win situation that might occur
in this change is that the material flow change triggers OLM to improve their business in general
and that OLM manages to successfully do this change. This would make OLM a more complete
and competitive supplier, giving them the opportunity to acquire new customers. Challenging
the suppliers is in accordance with SCM in order to improve the whole supply chain, however,
this might in turn affect Sandvik Mining due to that OLM´s capacity has to be shared with other
customers. This would thereby force Sandvik Mining financially due to that Sandvik Mining
might have to get a new supplier. We think that Sandvik Mining should consider the risk with
making OLM an ordinary supplier, is there a possibility that, maybe in a far or near future,
Sandvik Mining will lose OLM as a supplier together with their knowledge and capabilities.
What if neither company benefits from the new material flow? The worst scenario that
might occur is that neither of the companies benefits on the change. OLM is considered as a
rather important part of Sandvik Mining’s R&D, and by making OLM a more ordinary supplier it
could lead to that Sandvik Mining loses out on the expertise given by OLM in the product
development process. There is a strong relationship between long-‐term profitability and
investment in R&D according to Chiesa & Masella (1996), and in this case Sandvik Mining might
just be doing the opposite. We are of the opinion that it is important for Sandvik Mining to see
past the tangible cost savings that are presented in our material flow case and consider potential
consequences with OLM´s involvement in the development process if the material flow is
changed. We also think that by forcing OLM to deal with the new functions could affect OLM’s
core competence in a negative way, since it probably would be the key person at OLM that will
manage these new functions. In addition, OLM might not even be capable of managing new
functions. It is according to us a big challenge that OLM has to go through and OLM is the
company that has to deal with all the additional costs involved with the new material flow,
which can affect the company financially, making them the loser.
Is it even necessary to change the material flow? Maybe the relationship can be improved
just by structuring up the organization at Sandvik Mining. The risk with changing the material
flow is that the problems that actually exist at Sandvik Mining not are solved. In addition,
another problem might be created at OLM. The actual problem that seems to exist is that the
structure and boundaries towards OLM is poor, and not the actual purchasing of material or
transportation from the suppliers per se. At a first look there are some savings to achieve by
changing the material flow but at the same time the new costs have not been taken into
consideration in the business case due to its complexity and the risk is that the new costs will
exceed the savings. Furthermore, the risk of losing some of OLM´s core competence.
How does Sandvik Mining see the relationship with OLM in the future? Is the aim with the
relationship making OLM an ordinary supplier or to maintain the current competences and
capabilities? The real question maybe is if it is possible to change OLM to an ordinary supplier
and at the same time maintaining their core businesses.
67 |
Chalmers University of Technology | Conclusion
7 Conclusion
As stated in the purpose, this case study aims to clarify how the overall relationship looks
between Sandvik Mining and OLM. This was done by investigating and analyzing three areas that
were identified as improvement areas. These areas are the material flow, information flow and
processes that are performed regarding OLM.
The material flow business case was identified to have the highest potential for cost reduction
and lead time reduction. The information flow analysis showed that no formal structure exist
towards OLM and therefore the communication occurs in a unstructured way, causing priority
issues and extended and inaccurate lead times. Finally, the processes analysis of each role
involved with OLM showed that most tasks are made on experience, which makes the company
vulnerable in the future since almost nothing is written down on paper.
The first issue the case study cover is the material flow that is present between Sandvik Mining
and OLM. Currently Sandvik Mining is purchasing the material from raw material supplier,
receiving it at their plant before being sent to OLM. This material flow has been analyzed and a
new contingent material flow has been suggested and presented as a business case. In the new
material flow the material is re-‐directed from the raw material suppliers to go directly to OLM
and thereby skip the holdup at Sandvik Mining. The business case shows that Sandvik Mining
will reduce the lead times with 16 days, releasing approximately 1 370 000 SEK in tied up
capital. Furthermore, the business case also shows that there will be a cost saving of
approximately 570 000 SEK annually with the new material flow due to saving in the
transportation and holding cost.
The second issue the case study cover is the information flow between Sandvik Mining and OLM,
which today is vague. Furthermore, it turned out that the data about lead times is inaccurate and
no clear structure towards OLM exists. This has among other things led to priority issues for
OLM, which in turn has generated that the delivery precision is lower than other lego suppliers’.
The authors have recommended that Sandvik Mining should have two different approaches
towards OLM. When standard products are being discussed the only contact with OLM should
occur through the steering group that consists of the planner and the order processors. On the
other hand when new and test products are being discussed and developed it is of the essence to
have the communication between for instance the designers and OLM due to importance of
OLM´s involvement when developing new products. Therefore, the structure towards OLM for
new and test products should be the same as the current one, which was identified as a product
structure. However it must be emphasized that there are boundaries and that there is a steering
group which has the executive authority. Furthermore, the aims with the new structures are that
they should be applicable to be used for all lego suppliers that Sandvik Mining has.
The final area that the case study covers is the work processes of the different roles´ at Sandvik
Mining that regards OLM. All of the processes are illustrated and finally analyzed on a brief level.
The authors identified that the standardization of the processes are vague, which makes Sandvik
Mining vulnerable. Therefore, it has been recommended that the work processes has to be
standardized and written down in order to be able to improve the processes from the best
known practice and thereby make it easier to introduce new employees.
Finally, the thesis is wrapped up with a discussion about two important main questions that are
worth thinking through before performing any changes mentioned in the case study. The first
question is if the new material flow is the primary solution to the actual problem and the second
question is if the new material flow will be beneficial for all parties.
70 |
Chalmers University of Technology | Predicting Deviation in Supplier Lead Time and Truck Arrival Time Using Machine
Learning
A Data Mining Project at Volvo Group
MENG HUANG
MASOOD BAGHERI
Department of Technology Management and Economics
Division of Supply and Operations Managemen
Chalmers University of Technology
Abstract
The deviation in delivery performance from a company’s suppliers directly affects
the company’s performance, causing availability loss for the customer orders and
large costs for the rush transportation. If the deviation can be predicted in advance
and used as deviation alerts, actions can be taken in advance either to prevent the
deviation or decrease the impact of the deviation.
To predict the deviation in the supplier delivery performance from a buying com-
pany’s point of view, this thesis work specifically focuses on the first two phases of
a supply chain, namely supplier lead time from material suppliers and truck arrival
time from logistics service providers (LSP). In order to examine the possible imple-
mentation of machine learning, a data mining project has been conducted at Volvo
Group Service Market Logistics. The factors associated with deviation of supplier
lead time and truck arrival time are identified, while the corresponding features
are prepared under the constraint of the case company’s data availability. For pre-
dicting deviation in the two phases, two machine learning models are constructed
accordingly based on the characteristics of output and input features. The opportu-
nitiesandobstaclesalongthedataminingprocessinthecasecompanyareidentified.
The results show currently in the case company, both generated machine learning
models do not have enough predictive power in lead time deviation. This could
be caused by the absence of some key features that have strong associations with
deviation. However, the performance of the prediction model for truck arrival time
is regarded to be improved to a deployable level when the desired features are con-
structed into the model by the case company. Future recommendations regarding
constructing the desired features and improving the model performance are pro-
posed. In comparison, predicting deviation in material suppliers’ lead time could
be practical when the buying company get more information sharing from material
suppliers.
Keywords: Lead time deviation, Estimated time of arrival (ETA), Prediction, De-
livery precision, Machine learning, Supplier evaluation, Spare parts, Automotive.
v |
Chalmers University of Technology | 1
Introduction
In this chapter, the theoretical background and company background of this thesis
project is introduced, following by the aim of the project. The research questions are
thereby formulated and the scope of the project is presented.
1.1 Theoretical Background
Spare part supply chain is a high-margin business bringing in high profits for the
company. However, delivering spare parts is more complex than manufacturing the
products, since a spare part supply chain has to cover the aftermarket service for all
the products sold by the company. Customers also expect their things to be fixed
quickly when they break down, while their demands are intermittent because the
breakdown happens unexpectedly. These difficulties make only companies that pro-
vide the spare partefficiently can make revenues from aftermarkets (Cohen, Agrawal
and Agrawal, 2006).
The supply chain management in a company should match the demand and supply
(Jonsson,2008). Forecastingthedemandinordertomitigatetherisksofuncertainty
and availability loss of spare parts has received lots of research attention (Dekkeret
al., 2013). The uncertainties also come from supply sides (Heydari et al., 2009),
where deviation in lead time impacts the delivery precision and raises uncertainty
on the supply. According to Ioannou and Dimitriou (2012), lead time has direct
impacts on inventory and supply availability, and therefore the issue of managing
lead time has also been consistently discussed in the literature since the late 1960s.
To be specific, when a deviation occurs to the lead time, it results in the estimated
time of arrival (ETA) being not accurate and further disturbing inventory planning.
The inventory of spare parts is, therefore, going to fluctuate, causing stockouts when
spare parts arrive late or inventory holding costs when they arrive early (Heydari
et al., 2009). In particular, spare parts belong to maintenance inventories and the
stockout costs of the spare parts could be significantly high (Kennedy, Patterson
and Fredendall., 2002). Inspired by preventive and corrective maintenance (Mobley,
2002; Schmidt and Wang, 2018), if the deviation of lead time can be predicted be-
forehand, preventive actions can be adopted to minimize deviation, improving the
accuracy of ETA and secure delivery precision. Corrective actions can also be sched-
uled to mitigate the impacts of the deviation. For instance, to diminish deviation,
1 |
Chalmers University of Technology | 1. Introduction
more attention can be put on monitoring the supply process where it is predicted
to have deviated time of arrival and therefore the company can proactively take ac-
tions to avoid the deviation. To mitigate the impacts of deviated arrival time that
could bring fluctuated stock level, inventory planning can be updated considering
the deviation of ETA to ensure the availability of stock.
Overall, the successful prediction of the deviation on lead time can, firstly lead to a
lower total cost, because the right information of arrival time contributes to having
the right amount of spare parts in the inventory at the right time, saving both in-
ventory holding costs and inventory shortage costs (Carbonneau, Laframboise and
Vahidov, 2008). Secondly, it can improve customer satisfaction by securing their
vehicle up-time with the availability of spare parts needed in the warehouse (Car-
bonneau et al., 2008). Therefore, costs saving and capability of fulfilling customer
orders on time are the outputs of an accurate prediction of lead time deviation.
Since there are various companies cooperating in the supply chains, the performance
from supplier companies is going to affect buying companies’ performance. This is
the case especially for manufacturing industries including automotive, who relies
heavily on component suppliers (Krause, Handfield and Tyler, 2007). Therefore, it
is beneficial to predict the delivery performance from the buying companies’ per-
spective to secure their business operation.
Machine learning models are emerging to be used to predict suppliers’ performance
and predict the lead time or ETA in different transport modes, due to its ability to
capture the pattern from complex relationship between input features and output
performance (Witten et al., 2017). For example, predicting arrival time of truck
in distribution are discussed (van der Spoel, Amrit and van Hillegersberg, 2017).
Delay in passenger airplanes and freight trains (later than ETA) have also been
predicted using machine learning from transport handlers’ perspective (Belcastro et
al., 2016, Takacs 2014, Barbour et al, 2018). However, for material supplier per-
formance, existing literature only predicts supplier overall performance rather than
specifically focusing on delivery precision (Jiang et al., 2013; Khaldi et al., 2017).
For the transportation, the performance of prediction models varies with different
input features. So far, we have not found literature that is based on input variables
of organisation and human to predict truck arrival time with machine learning.
1.2 Company Background
Volvo Group (Volvo) Service Market Logistics (SML), as one of the departments in
the case company where this project is performed, is responsible for the develop-
ment and optimization of the spare part supply chain which strives for securing the
availability of spare parts at the lowest possible costs.
2 |
Chalmers University of Technology | 1. Introduction
To achieve this goal, the target of delivery precision performance from logistics ser-
vice providers (LSP) in SML is 97%. It means 97% of transportation delivery shall
not arrive late on each node. However, due to the fact that lead times are negotiated
with their suppliers and set in the planning system for a longer period of time since
the cooperation starts and there are various uncertainties in supply process, the de-
viation occurs frequently in lead time. For the spare parts of Volvo truck in Europe
in 2018, around 37% delivery does not meet the ETA at their central distribution
center (CDC) according to predefined transportation lead time (TLT). Among the
deviation, 27% of them arrived earlier and 10% of them arrived later than ETA. Pre-
vious than that transportation delivery goal, the target for the material suppliers’
delivery precision is 95%, which means 95% of the orders from material suppliers
shall not be ready later than scheduled. However, for the previous performance in
2017 and 2018, merely 77% of them does not have deviation in supplier lead time
(SLT) and was dispatched on time, with 9% of them dispatched earlier than sched-
uled, and the remaining 14% dispatched later than estimated.
This big share of deviation could directly bring fluctuation in inventories. Spare
parts arriving earlier than estimated are bringing extra tied-up capital, inventory
costs and disturbing the work schedule in warehouses, while late-arrived spare parts
could either cause extra delivery costs in recovering the back-orders by expediting
logistics using air transport, or become excess inventory and end up being scrapped
because of missing out to supply the demand. As it is important for Volvo to fulfil
customers’ demand at a lower cost, there is a need for predicting lead time deviation
for monitoring the delivery precision performance on their material suppliers and
logistics service providers (LSP) in order to proactively checking ETA of spare parts
and take actions.
In Volvo, the importance of big data is increasingly raising attention. More and
more data are collected and analyzed. These new data resources combined with
advanced analytic methods are creating new opportunities to reap the fruits of data
mining to benefit business. Volvo has realized the power of machine learning models
in prediction and has been initiating data mining projects to explore its possible us-
ages and potential benefits. Therefore, this study targets on predicting deviation of
lead time on its suppliers of material and transportation by implementing machine
learning.
1.3 Aim
The aim of this thesis is to evaluate whether and how machine learning modelling
can be implemented to predict lead time deviation from buying companies’ suppliers
of material and logistics, under the consideration of achieving benefits of a predic-
tion model in the current stage of the case company Volvo SML.
3 |
Chalmers University of Technology | 1. Introduction
To achieve the aim of implementing machine learning models to predict lead time
deviation, the first research question is to investigate how the company can utilize
the lead time prediction. This question sets the business goal and answers potential
benefits of this data mining project.
RQ 1: What are the benefits of predicting lead time deviation for buying com-
panies?
The second research question is to investigate the factors that are associated with
deviation from the buying company’s perspective. These factors are the basis for
features construction for machine learning modelling.
RQ 2: What are the factors that could be associated with lead time deviation
perceived by buying companies?
However, only the factors that can be represented with available data in the com-
pany’s database can be analyzed and constructed into the prediction model. This
research question reflects the limitation existing in the case company for the con-
struction of the model and contributes to set the data mining goal of this project.
RQ 3: Which data are available to be used as features when building the pre-
diction model of lead time deviation at Volvo SML?
The fourth question is to develop a prediction model by testing different machine
learning strategies and algorithms. The modelling process is based on Volvo’s situa-
tion considering the benefits that the company can practically achieve in the current
stage. Theresultsofmodellingwillbealsoexaminedandinterpretedregardingtheir
usability.
RQ 4: How should the prediction model be built using machine learning con-
sidering the practicality of use in the current stage at Volvo SML?
1.4 Scope
In order to fulfil the aim of this thesis project, a certain scope is needed. The scope
of the thesis is focusing on the spare parts that belong to Volvo truck in European
region. Further, for the scope of lead time, the chosen phase will be examined from
the moment that Volvo places orders to its material suppliers and shipped by LSP
until they arrive at the CDC in Ghent, Belgium. The reason for choosing this in-
bound flow is because it currently suffers from the largest deviation and this flow
is at the beginning of the supply chain which has cascade effects on later processes.
In this project, this lead time is named inbound flow lead time and it consists of
two phases which are supplier lead time (SLT), and inbound transportation lead
time (TLT). The SLT is the time taken by the material suppliers to get ready for
4 |
Chalmers University of Technology | 2
Literature Review
In order to support the analysis and discussion by providing theoretical resources
and domain knowledge for machine learning, a literature review is conducted in this
chapter. It is divided into two parts with the first part reviewing spare parts supply
chain, previous work and current state of predicting lead time deviation, while the
second part including the last two sections is introducing machine learning.
2.1 Frame of Reference
This section introduces the frame of reference which helps to present the context
of spare part logistics and the application of machine learning in the area of sup-
plier evaluation and ETA prediction. They are corresponding to the subjects of this
project.
2.1.1 Spare Part Logistics Context
The requirements for planning spare parts logistics are different from the logistics
of other material from several aspects (Huiskonen, 2001). Firstly, the service re-
quirement of logistics is high due to the remarkable costs and penalties for spare
parts shortage. However, the demand for spare parts is sporadic and hard to predict
which bring high risks of late delivery. Secondly, due to the decrease of the buffers
of time and material in the supply chain and production systems, streamlining the
spare parts logistics is under the pressure (Huiskonen, 2001).
Mostpapersareaddressingtheserequirementsbyfocusingontheinventorymanage-
ment of spare part locally rather than considering the whole supply chain (Zanjani
& Nourelfath, 2014). However, inventory optimization often has strict assumptions
and difficult to apply. There is a need to increase the collaboration between different
actors to plan spare parts logistics to deal with the special requirements of spare
part logistics (Huiskonen, 2001).
7 |
Chalmers University of Technology | 2. Literature Review
2.1.2 Supplier Evaluation
One aspect of collaboration for today’s supply chain management is to maintain a
long term relationship with suppliers by having a fewer number of suppliers with
reliable performance. Hence, it is important to evaluate the suppliers’ performance
effectively in order to maintain the right suppliers (Ho, Xu and Dey, 2010). Since
automotive companies are especially dependant on their sub-component suppliers,
their performance is much affected by their supplier performance in delivery time,
reliability and flexibility, according to Krause, Handfield and Tyler (2007). It means
if a supplier improves its production time then its industrial customers could get
their order faster as a consequence. Therefore, to evaluate the performance of their
suppliers is very important for buying companies’ performance.
As a multiple criteria decision-making problem, supplier evaluation can have several
quantitative and qualitative criteria. The relationship between these criteria and
supplier performance could be complex (Rezaei, Fahim, and Tavasszy, 2014). While
existing papers mainly discuss supplier evaluation for the purpose of choosing the
right supplier, which belongs to a pre-evaluation at a strategy level, very few pa-
pers are focusing on adopting post-evaluation at an operational level ( (Khaldi et
al., 2017). Only Khaldi et al. (2017) adopt artificial neural network algorithm to
evaluate and predict the hospital’s suppliers performance from their transactional
contracts and paperwork of delivery articles including delivery delays, the number of
partial deliveries, turnovers, amount of orders. The output of the prediction model
is the efficiency score of suppliers. Jiang et al. (2013) conduct an experiment to
forecast new suppliers’ classification in terms of their performance and efficiency.
They train the support vector machine model with the input of cost reduction per-
formance, price, delivery, quality.
For predicting supplier’s lead time deviation, in essentials, it is a supplier evaluation
taskwhich focusesspecifically onsuppliers’deliveryprecision performance. Delivery
precision or delivery reliability refers to the ability to delivery according to schedules
or promises (Sarmiento et al., 2007). The higher the delivery precision, the lower
the deviation of lead time. This research has not been performed previously to our
best knowledge.
2.1.3 ETA/Lead Time Prediction
For TLT prediction, there are literatures developed in each transportation scenario,
such as train, road and flight. However, according to a literature review conducted
by Van der Spoel, Amrit, and Hillegersberg (2017), there is very few literature pre-
dicting arrival time focusing on trucks. Therefore, this study considers to learn from
the practice from each mode of transportation, one up-to-date paper is chosen and
described for a review and summarized into Table 2.1.
8 |
Chalmers University of Technology | 2. Literature Review
Van der Spoel, Amrit, and Hillegersberg (2017) state that unlike the travel time
which may be well predicted by using weather and traffic information, the truck
arrival time could be much affected by human and organizational factors such as
planning departure time. That means there is the difference between predicting lead
time and arrival time. The result of lead time prediction cannot be directly applied
to arrival time prediction without considering planning departure time. They test
it by predicting arrival time only using those weather and traffic information. The
response output is classified by the tardiness of trucks arriving at the distribution
center. The classes are roughly from very early and slightly early to very late and
slightly late. They test a set of algorithms such as random forest. Finally, the result
is as estimated. The prediction power of the developed models for arrival time is
not satisfying since human and organization factors are not included as features.
Belcastro et al. (2016) predict flight delays by focusing on weather condition since
the weather is the cause of delay for more than 1/3 of the flights. They have high
precision and recall score up to 86% for a large delay threshold to be 60 minutes.
The threshold means when a flight arrives more than one hour later than the ETA,
this flight is counted as ‘late’.
Barbour et al. (2018) predict the travel time of a freight train in real time in order
to generate ETA. A full network state information from transportation handler in-
cluding physical train characteristics and train crew information are the input for
having regression results. Compared to the current analytical method calculating
the travel time only considering the network topology and traffic through particular
routes, they manage to improve the performance by over 60% using random forest.
Table 2.1: Review of predicting ETA/machine learning with machine learning
Author(s) Subject Classifi- Input data Model Remark
cation/
Regression
Van der Truck ar- Classifi- Traffic informa- M1 Low prediction
Spoel rival time at cation tion, Weather ensemble, power 72% ac-
et al., (2017) Distribution information Random curacy
center Forest...
Belcastro et Flight delays Classifi- Weather Condition MapReduce Accuracy 85.8%
al.(2016) cation Flight information Recall 86.9%
Barbour et FreightTrain Regression Afullnetworkstate Random maximum
al. (2018) Arrival Time including physical forest, predictive im-
(travel time) train character- Support provements of
istics, train crew vector re- over 60% using
information gression, random forest
Neural compared to the
network current method
9 |
Chalmers University of Technology | 2. Literature Review
2.1.4 Conclusion from Frame of Reference
From the frame of reference, we can conclude that implementing machine learning
model on predicting suppliers’ delivery precision is an unexplored topic. Existing
literature only implements machine learning to predict the overall performance of
suppliers based on multi-criteria. Therefore, it remains to explore whether supplier
delivery precision can be predicted with machine learning models from the buying
companies.
Similarly, plenty of work has been done on predicting ETA for various transporta-
tion modes but few of them focuses on truck. For flight delay prediction, since the
weather is one of the major causes for the delay, only considering weather and flight
informationcouldgenerateagoodpredictionresultwithmachinelearning. However,
for predicting ETA of the truck, only considering weather and traffic information
is not enough to have good prediction power since organization and human factors
could frequently cause deviation in arrival time. When a full network state infor-
mation including human and organization factors is used for predicting ETA of the
freight train, a significant improvement of prediction is made compared to the pre-
vious prediction model where only traffic and route information is used. Therefore,
our work will try to consider organization and human factors into the prediction
model for ETA of trucks, since it is unexplored which information could be effective
tobeusedasinputfeaturesformachinelearningmodelstopredictdeliveryprecision
of LSP.
2.2 Machine Learning Tool and Terminology
This section is going to introduce machine learning and its relevant terminology
such as input and output, algorithm selection, classification and regression models,
boostingandbagging, randomforest, catboostandgradientboosting, handlingclass
imbalance.
2.2.1 Fundamental Machine Learning Definition
Machine learning is a field covering the main techniques used for data mining which
is finding the patterns in the substantial amount of data. The discovered patterns
must be insightful which can assist decision making (Witten et al., 2017). There
are two extremes about a pattern, from a black box whose mechanisms are incom-
prehensible to a transparent box whose construction reflects the formation of the
pattern. The difference between them is whether the patterns can be explained and
interpreted. Both of them could lead to good predictions and knowing the inputs
and outputs are way more important than understand the mechanisms in between
(Witten et al., 2017).
10 |
Chalmers University of Technology | 2. Literature Review
There are some fundamental machine learning definitions. Input is including con-
cepts, instances and feature. Concept is the thing to be learned. The input to a
machine learning model is a set of instances that needs to be classified, associated or
clustered. Each instance is an independent example of the concept used for learning
or evaluation. There are features which is another set of predefined attributes that
are measuring various aspects of the instance (Witten et al., 2017). Dimension of
features measures the number of features.
There are typically two types of features for machine learning, namely categorical
and continuous one. According to Prokhorenkova et al. (2018), categorical features
refer to a discrete set of values that are incomparable to each other in a numeri-
cal way. The measurement scale of the categorical features consists of a different
set of categories (Agresti, 2018). Categorising the features can be implemented in
three different ways. The simplest one is regarded to the situations of having bi-
nary features when the values could be categorised in “0” or “1” or “YES” or “NO”
segments. Furthermore, the categorical features could be mapped on an ordinal
scale. For instance, they could be classified such as: “very late”, “late”, “on time”,
“early” and “very early”. These features are also called “ordinal variables”. Nominal
features are the final segment according to Agresti (2014). Nominal features have
no numeric values and are independent of each other. These features are normally
used to identify something (e.g. countries) and have not any kind of natural order.
In contrast, continuous features are referred to as the variables that have an infinite
number of possible values. Label is the values or categories belonging to instances
(Mohri, Rostamizadeh and Talwalkar, 2012).
The input instances are divided into training set and test set. Training set is used
to train a machine learning model, while the test set is used to evaluate the perfor-
mance of the model. The test set is separated with the training set and not available
at the training phase. The output of the model is the form of prediction on new
instances (Mohri et al., 2012).
2.2.2 Algorithms and Feature Selection
Knowing which algorithm is likely to deliver a good performance for the investigated
problem is known as an algorithm selection problem (Rice, 1976). There is no uni-
versally best algorithm for solving a vast problem domain (Wolpert and Macready,
1999). Identifying the most suitable machine learning algorithms which can discover
the relationship between the output and the relevant features is a challenging issue
(Lingitz et al., 2018). It is necessary to well understand the characteristics of the
problem in order to choose the suitable algorithms (Smith-Miles, 2009).
There is the ensemble method which can adopt multiple machine learning algo-
rithm to achieve better predictive performance. Based on the different strategy, it
is categorized into boosting and bagging. García-Pedrajas et al. (2012) describe
the function of boosting by saying that it builds an ensemble in a step-wise manner
11 |
Chalmers University of Technology | 2. Literature Review
by making a new classifier and add it to the ensemble. The logic of this process is
that the new classifier would be trained towards the biased samples. If any sample
has been misclassified during the boosting process they will be assigned by a higher
weighted value (García-Pedrajas et al, 2012). Boosting is a general method to use
in order to improve learning algorithms since it is capable to reduce the errors of
weak learning algorithm (Freund and Schapire, 1996).
In terms of the bagging method, it is a set of predictors based on bootstrapped
aggregated samples in order to achieve an aggregated performance (Breiman, 1996).
For predicting specific classes, the majority of the votes from multiple predictors for
one class would be selected. For the prediction of a numerical output, the average
value of the output from the aggregated predictors would be considered.
When adopted machine learning, the first decision is to choose between supervised
machine learning which assumes that training examples are labelled, unsupervised
machinelearningwhichhasfocusedontheanalysisofunclassifiedexamples, orother
techniques such as semi-supervised machine learning or reinforcement learning (Lin-
gitz et al., 2018). Semi-supervised learning consider both labeled and unlabelled
data which is commonly used when some labeled data are expensive to obtain but
unlabeled data could also help achieve better model performance. Reinforcement
learning is intermixing the training and test phase, for each move receive immediate
rewards to help prediction(Mohri et al., 2012). According to Öztürk et al (2006)
supervised learning is considering the relationship between the output and the in-
dependent or explanatory features in a model. It aims to predict output based on
input features with a prerequisite of a known training set (Pfeiffer et al., 2015).
Feature selection is another key process in machine learning. There are many possi-
ble benefits with feature selection: decreasing dimensions for improving prediction
performance, providing faster and effective predictors with lost cost, assisting to
understand the underlying process of data generation (Guyon and Elisseeff, 2003).
According to Dash and Liu (1997), in real word practice, most classification prob-
lems require the supervised learning with each instance associated with a class label.
Since the relevant features could not be known beforehand, the candidate features
are often selected for their representativeness for the domain. Unfortunately, many
of these candidate features are often irrelevant or redundant to the output concept
and not affecting the output result. However, as soon as the size of features or
dataset are up to hundreds to thousands, reducing them could significantly increase
the speed of machine learning (Dash and Liu, 1997; Guyon and Elisseeff, 2003)
2.2.3 Classification and Regression models
Classification and regression are two important data mining missions for supervised
machine learning. Both of them contribute to building a data-driven model to learn
an unknown underlying function that illustrates the relationship between several
input features and one target variable as the output of the function (Cortez and
12 |
Chalmers University of Technology | 2. Literature Review
Embrechts, 2013; Lingitz et al, 2018). To compare the regression and classification
model, this selection should be based on predictive capability, computational re-
quirements and explanatory power (Cortez & Embrechts, 2012).
The difference between these two types is made by the existence of categorical and
continuous features in a model. When the output in a predictive model is set to be
categorical variables then the classification techniques would be used. In the case of
having a output in the form of a continuous value, the regression techniques would
be applied (James et al., 2013).
2.2.4 Random Forest
Random forest has combined two powerful algorithms namely bagging and ran-
dom feature selection (Breiman,2001; González et al., 2014). According to Breiman
(2001), random forest is an ensemble Classification and Regression Trees (CART)
classifiers, that each decision tree is created without any pruning and bagging algo-
rithm is applied in order to create a “forest” of classifiers voting for specific labels.
Each tree is considered as a predictor. Random forest could be used for both clas-
sifications and regression problems. Pfeiffer et al. (2015) adopt the random forest
regression to estimate the lead time as a continuous output variable. They argue
the random forest model has better performance than the decision tree model and
multiple linear regression model. According to González et al (2014), random forest
is capable to capture the complex interactions with different data structure and it
is also robust to over-fitting problems.
2.2.5 Gradient Boosting and Categorical Boosting
Gradient boosting has been used as an advanced machine learning technique for
many years, which can handle complex data sets in an effective way. According to
Zhang & Haghani (2015), gradient boosting is a regression tree based algorithm that
builds a model in a stage-wise fashion and updates it by minimizing the expected
valueofcertainlossfunction. Gradientboostingbasicallyappliesgradientdescentin
a functional space to build ensemble predictors. Friedman (2001) describe gradient
boosting as an algorithm that is highly robust and explainable for both regression
and classification problems.
According to Prokhorenkova et al. (2018), categorical boosting(Catboost) is the
execution of gradient boosting that uses binary decision trees as base predictors. In
Catboost, the decision trees have the same split criterion along with the entire level
of the trees. These trees are less prone to over-fitting and have a higher speed of
processing time for the testing data set. Prokhorenkova et al. (2018) claim that
Catboost outperformed the other advanced gradient boosting algorithms, XGBoost
and LightGBM on plenty of different machine learning tasks. Dorogush, Ershov
and Gulin (2018) introduce Catboost as an algorithm that has been successful in
13 |
Chalmers University of Technology | 2. Literature Review
dealing with categorical features which are in practice very hard to deal with. The
authors also mention that Catboost algorithms can handle the over-fitting problem
in a convenient manner.
2.2.6 Handling Class Imbalance
Handling class imbalance distribution is a significant topic happening frequently in
practice. Class imbalance arises when classes are represented unequally. Namely,
most of instances are labelled as one class, while the rare instances are labelled as
the other class which might be of more interest or importance. It is crucial that a
classification model should be able to achieve higher identification capability on the
rare occurrences in datasets. Many traditional classifiers are not compatible with
the learning task with imbalanced classes (Kotsiantis, Kanellopoulos and Pintelas,
2006). According to Ali, Shamsuddin and Ralescu (2015), there are two problems in
handling class imbalance. One of the main concerns is that data mining performers
could be accuracy driven. The traditional way of examining a model performance
focus on accurate performance. Classification algorithms selected for their high
accuracy performance are likely to group all the data into the majority class to min-
imize overall error. This is often at a cost of misclassifying the rare instances. In
a class imbalance dataset, classification accuracy tells very little about the minor-
ity class and may give a misleading evaluation of classifier performance. Another
issue in learning with class imbalance distribution is that standard classification al-
gorithms are based on the assumption of the evenly distributed underlying training
set. Failing to consider the skewed distribution of data is most likely to hinder the
classification performance (Ali, Shamsuddin and Ralescu, 2015).
The classification performance for imbalanced data is also subjective to the size of
the dataset (Kotsiantis, Kanellopoulos and Pintelas, 2006). It may be even worse
for an small imbalanced dataset compared to the larger one, due to the insufficient
sample size of instances representing minority class for learning. On the contrary,
the effects are relatively less severe with larger datasets, as the minority class is bet-
ter represented by a larger size of examples (Kotsiantis, Kanellopoulos and Pintelas,
2006).
To handle class imbalance classification, sampling techniques and cost-sensitive
learning are commonly applied. Sampling techniques are used to either remove
a small number of examples from majority class or over-sample minority class or
both. By introducing this sampling step, the discrepancy between the two classes is
minimized so that traditional classification algorithms can work well. For example,
Balanced Random Forest, incorporating under-sampling majority class technique
and the ensemble learning, artificially re-balances the class distribution to ensure
that classes are equally represented in each tree (Chen and Breiman, 2004).
14 |
Chalmers University of Technology | 2. Literature Review
Cost-sensitive learning approaches, on the other hand, impose an expensive cost
on a classifier when a misclassification happens in order to emphasize any correct
classification or misclassification regarding the minority class (Kotsiantis, Kanel-
lopoulos and Pintelas, 2006). For instance, in Boosting algorithms, different weights
are placed on the training distribution in each iteration. In order to emphasize
misclassified examples in the next iteration, boosting increases the weights on the
misclassified examples and decreases the weights on the correctly classified examples
after each iteration. Since minority classes are more likely to be improperly clas-
sified in comparison with majority classes, boosting may improve the classification
performance through increasing the weights of the examples from rare classes. Also,
as boosting effectively rebalance the distribution of the training data, it can also
be considered as an advanced sampling technique (Kotsiantis, Kanellopoulos and
Pintelas, 2006).
2.3 Evaluation Metrics for the Prediction Models
Since the overall accuracy could insufficiently or even misleadingly evaluate a clas-
sifier performance (Visa, 2006; Japkowicz and Stephen, 2002; Wang and Mendel,
1992), the confusion matrix and its derivations are introduced as a more proper
way to summaries the performance results. Feature importance is also introduced
as another measurement for the input features.
2.3.1 Confusion Matrix
AconfusionmatrixshowninTable2.2. istypicalforevaluatingthemachinelearning
models’ performance with imbalanced classes. Class “C” is regarded as the minority
class which is in the focus, while “NC” is a combination of all the other classes.
There could be four kinds of results when detecting class “C” (Chawla et al., 2003).
The first one is true positives which correctly recognized focused class examples.
True negatives are those correctly identified examples that do not belong to the
focused class. The third factor, false positives, considers the examples that were
incorrectly assigned to the focused class and finally the last one is false negatives
which were not successfully recognized as focused class examples. These four factors
constitute a confusion matrix (Chawla et al., 2003).
Table 2.2: Confusion matrix defines four possible scenarios when classifying class
“C” (Chawla et al., 2003)
Predicted Class “C” Predicted Class “NC”
Actual class “C” True Positives (TP) False Negatives (FN)
Actual class “NC” False Positives (FP) True Negatives (TN)
15 |
Chalmers University of Technology | 2. Literature Review
From Table 2.2., recall, precision and F-value are defined as follows:
Precision = TP/(TP +FP) (2.1)
Recall = TP/(TP +FN) (2.2)
(1+β2)∗Recall∗Precision
F −value = (2.3)
β2 ∗Recall+Precision
The performance metrics derived from the confusion matrix are including precision,
recall, F1 score which comprise of a classification report for the modelling result.
Precision measures the exactness, which is the proportion of correctly predicting
classes. It shows the ability of a classifier for avoiding misclassifying negative classes
into the positive class. Recall measures the completeness, which represents a clas-
sifier’s ability to learn positive class. It is calculated by the proportion of correct
detection of positive example out of all positive example in the data. F-score is a
way of balancing the measurement between precision and recall. As the β is com-
monly set to 1, therefore F1 score is used for classification (Sokolova and Lapalme,
2009).
The common pursue of all learning model is to improve the recall while not to sac-
rifice the precision. However, there is often the conflicts between them and it may
be difficult to improve both of them at the same time. This situation is especially
true when one or more classes are rare (Chawla et al., 2003).
2.3.2 Feature Importance
The increasing popularity of machine learning models is largely credited to their
capability to handle high-dimension data with large number of predictors and other
advantages including relatively good accuracy, robustness, ease of use (Breiman,
2001). However, it is common that not all the features are important and some of
input features can be relative irrelevant or redundant in data mining. Identifying
the most important features is beneficial because it indicates which features have
the highest predictive power for the model and may help the domain users to have a
better understanding of the problem. It can also help to develop recommendations
for the future, and it may lead to changing the role of the underestimated features
more seriously (Petkovic et al, 2016). To identify the features with the most signif-
icant impacts on predictions, feature importance is one of the most commonly used
measurements, which facilitates feature selection and model interpretation.
The most widely used feature importance measures are the impurity importance
and the permutation importance (Breiman, 2001). The impurity importance, also
known as Gini importance, is based on the mechanism of mean decrease of impu-
rity. It is the default feature importance measure embedded in some most popular
implementation platform such as R and scikit-learn in Python. In the impurity
importance, a feature is considered as important if it is effective at diminishing
uncertainty for classifiers or variance for regressors. The impurity importance for
16 |
Chalmers University of Technology | 2. Literature Review
a feature in random forests, for example, is computed by adding up all impurity
decrease measures of all nodes in the forest where a split on this feature has been
made, normalized by the number of trees. Another type of importance measure, the
permutation importance is also known as mean decrease of accuracy. Under this
mechanism, the important features are those positively contributing to reduce the
prediction error.
Despite its popularity, for years, the impurity importance is acknowledged to be
biased. The impurity importance is likely to inflate the importance of categorical
variableswithmanycategoriesandcontinuousvariables(Breimanetal.,1984; Strobl
et al., 2007), also in favor of variables with high category frequencies (Nicodemus,
2011). The permutation importance, on the other hand, is safe from these concerns
(Nicodemus et al., 2010; Szymczak et al., 2016; Ziegler and Konig, 2014). However,
the permutation importance can be extremely computationally intensive when en-
countering high dimensional data. Also, Calle and Urrea (2011) argued that feature
importance rankings based on the impurity importance can be more robust over
those obtained with the permutation importance.
17 |
Chalmers University of Technology | 3
Methods
In this chapter, the methods that were used to conduct this project are described.
First, the literature review was then conducted and also throughout the entire process
of the project. Then the CRISP-DM (Cross Industry Standard Process for Data
Mining) methodology and the reasons for choosing it are introduced. The correlation
between methods and research questions is also demonstrated. Finally, the reliability
and validity issues are described in the end section.
3.1 Literature Review
There are several reasons for conducting a literature review at the beginning of and
throughout the project. Bryman and Bell (2015) describe the first thing is to be
aware of and understand what has been already discussed in the research area. Sec-
ondly, it also gives a way for authors to develop an argument about demonstrating
the significance of the project and where it contributes. Beyond that, a literature
review with an interpretation from reliable sources in the research field could also
increase the credibility of the project. Based on the above reasons, a literature
review was conducted with the purpose of providing information for four research
questions and assisting the data mining process for realizing the aim of the project.
We searched literature from electronic database including Scopus, Google scholar
andChalmersLibrary. Thekeywordsusedinthesearchincludingthecombinationof
lead time deviation, estimated time of arrival (ETA), prediction, delivery precision,
machine learning, supplier evaluation, automotive. Peer reviewed articles and books
were examined and used in the literature review.The result of the literature review
is compiled in the chapter 2.
3.2 General Strategy and Process
The most commonly used process for data mining projects is CRISP-DM (Marban,
Mariscal and Segovia, 2009) It is process model being developed by a group of data
mining leaders for carrying out data mining projects. The purpose of this process
model is to make these projects more reliable and replicable with less money and
time spent (Wirth and Hipp, 2000). Wirth and Hipp (2000) discuss that the pro-
cess can not only be performed by experts, but the novices with less experience and
19 |
Chalmers University of Technology | 3. Methods
technical skills can benefit in a limited time. This is due to the characteristics of
CRISP-DM being both structural and flexible depending on whether it is generic or
specialized process. For less experienced people such as master students, we can get
guidance and structure of the project, as well as advice for each process.
TheprocessesofCRISP-DMfromgenerictospecificaredescribedasPhases,Generic
Tasks, Specialized Tasks and Process Instances (CRISP-DM, 1999). For the top
level, the phases of the model include business understanding, data understanding,
data preparation, modelling, evaluation and deployment representing the life cycle
of a data mining project. The second level is generic tasks with its intention to cover
all data mining situations. The third level aims to describe what actions should be
taken within the general tasks. The fourth level is a requirement of recording the
actions, decisions and results during the process.
We adapted the generic CRISP-DM process based on our data modelling project,
and the process is summarized in Figure 3.1. There are six phases in the CRISP-DM
process that are described in the following sections.
Business Data
Understanding Understanding
Collect Initial Data
Determine Business Objectives
Describe Data
Assess Situation
Verify Data Quality
Determine Data Mining Goals
Select Data
Integrate Data
Data Preparation
Deployment Clean Data
Construct Data
Generate Test Design
Select Modelling Technique
Evaluate Result
Build Model
Review Process
Assess Model
Determine next steps
Evaluation Modelling
Figure 3.1: Illustration of the data mining process based on CRISP-DM (1999)
3.3 Business Understanding
The first phase is about understanding the business. Business understanding in-
volves figuring out the feasible goals based on the situation and requirements from
the business perspective to achieve potential benefits. Therefore, qualitative data
about business were collected by means of conducting interviews and examining
internal documents in the company in order to set a feasible goal.
20 |
Chalmers University of Technology | 3. Methods
3.3.2 Internal Documents
For a case study research, the most crucial value of internal documents is to au-
thenticate and argue the evidence from other sources (Yin, 2018). With the inter-
nal documents, this project gained up-to-date knowledge about the case company’s
structures and business processes. The collected information also became evidence
to support arguments from interviews. The internal documents used in this project
were found in the internal database of the case company, including company pre-
sentation, process description system and the team places. These documents could
exist in the form of PowerPoints, word documents and other informative data from
databases.
After this stage, the goal of the business was defined to respond to the research ques-
tion 1. To answer the research question 2 about the factors of lead time deviation,
a pile of factors were compiled after conducting the literature review, interviews
and examining internal documents. A list of preliminary potential features was also
identified in this process.
3.4 Data Collection and Understanding
Forthedataunderstandingprocess, oneinvestigatedaspectwastocollectthehistor-
ical data of lead time deviation performance, which was used as the output variable
for modelling. Another aspect was gathering those available data that could asso-
ciate with factors of the deviation of lead time identified in the first stage. These
quantitative data were extracted from different databases in the case company as
archival records, as Table 3.2 shows. Historical lead time performance data of sup-
plier lead time was extracted from the Business Intelligence where the previous two
years data (2017 and 2018) were included. The data related to features of the first
model were also extracted from business intelligence and the reports generated from
supplier management portal VSIB. For the second model, most of the data were
extracted from the logistics management portal Atlas. These data were limited to
the previous one rolling year as the maximum amount of data the system held at the
time the project was conducted. Noted from the transportation delivery precision
report, there is up to 30% of delivery where goods were not delivered according to
planned deliveries. These missing deliveries were deleted and not considered into
the calculation of delivery performance since they are not generated the output of
delivery whether they are on time or deviated.
Then data understanding was to get to know the data about its variability and
availability, including the quality and quantity of the data. Since the business goal
needs to be translated into the goal of data mining, the availability of the data in
the company was under consideration. Hence the data mining goal was developed.
22 |
Chalmers University of Technology | 3. Methods
Table 3.2: Main data source from the case company’s database
Phase Data Sources Content
Historical lead time Business Intelligence – Parts- 2017&2018:
performance (Output DWH ver1.5 – For Std Report 402,708 pieces
Supplier
Variable) Developer of records
Lead
Time Features Parts: Business Intelligence – Segmentation,
PartsDWH ver1.5 Sales level spend,
Suppliers: VSIB – supplier man- delivery precision,
agement portal ...
Historical lead time LSPs portal Atlas 2018.04-2019.03:
performance (Output Filter: all volvo truck parts were 49,948 pieces of
Inbound
Variable) ended in CDC Ghent records
Trans-
portation Features Parts: Business Intelligence – Weight, volume,
Lead PartsDWH ver1.5 country ...
Time Suppliers: LSPs portal Atlas
Consignors and Volvo logistics
scheduling: Atlas
3.4.1 Delimitation in the Data Collection
There were a few limitations in the data collection phase. Firstly, when sampling
data from the data warehouse, the period was limited to what the data warehouses
hold. For the transportation phase, the data are recorded for one rolling year.
Therefore the amount of data for training were limited to one year period, which
could bring problems of bias and robustness.
The evaluation results of material suppliers were extracted from the supplier man-
agement system VISB. The options for evaluation period are from past three months
to past one year, the granularity of the evaluation results such as delivery precision
is limited by being made as average value for that chosen period.
There were data related to factors that were scattered in lots of separate reports
but not integrated into the data warehouses. In this sense, these data were not able
to be gathered and used as features for modelling. For example, the logistics audit
results of LSP exist in individual excel files for each LSP, then these data were not
utilized as a potential feature.
There were factors that relate to deviation but suffering from the data quality in
the system and not being used as a feature. For example, the departure time of
truck could have effects on deviation since it affects the arrival time of a truck to a
warehouse which could cease operation during the night and the late arrival truck
need to wait for one night to be processed. However, the departure time is not
precisely recorded in the system and therefore not suitable to be used.
Therewasthedatatransparencyissuethatthenamesofsomeitemsinthedatabases
were confusing without further explanation. In order to make sure the right data
23 |
Chalmers University of Technology | 3. Methods
was used, it also took time for us the data practitioners to find who can explain the
data in the company.
Sensitive information such as the relationship between suppliers delivery perfor-
mance and evaluators in the buying company was also not gathered and examined.
3.5 Data Preparation
The data preparation phrase is including all the actions that creating a final data
set which were fed into the modelling from the raw data including selecting data,
cleaning data, constructing data, integrating and formatting data.
3.5.1 Transferring Categorical Variable
There were many categorical variables in the feature list, in order to quantify them
and feed them into modelling, a function called dummies in the commonly used
python package Pandas was used to turn a categorical variable into a series of zeros
and one. One example is illustrated below, the feature of categorical variable ‘stack-
able’ is divided into two columns with ‘1’ represent of the characteristics being true,
and ‘0’ for not being true.
Stackable Stackable Non-stackable
Yes Get dummies 1 0
No 0 1
... ... ...
Figure 3.2: Transferring categorical variables into dummy variables
However, some categorical variables have a lot of classes such as 59 kinds of seg-
mentation of spare parts. When directly getting dummies for these variables, the
input data will get lots of columns with each one having little weight. Therefore,
these categorical variables were reduced into a reasonable amount of columns by
reconstructing and combining them based on some criteria. Segmentation of Volvo
spare parts is a comprehensive measurement defined in terms of criticality, life cycle,
cost and order frequency. For segmentation result, there are five different initial let-
ters from ‘A’ to ‘E’ as main catalogues. From ‘A’ to ‘D’, they represent four kinds
of criticality code, and ‘E’ represents non-critical parts. The criticality of a part
depends on specific function groups and vital codes. Under each letter, there is the
second letter starting from ‘A’ to ‘L’ for the sub catalogues representing the cost,
life cycle and order frequency information. Vital code, cost, order frequency are
available as independent features, while using function groups directly may result
too many categories, and life cycle phase is not directly available. Segmentation was
24 |
Chalmers University of Technology | 3. Methods
adapted to present information of function groups by keeping the main catalogues,
andclusteredthesecondcataloguesinto‘fast’and‘slow’toroughlyrepresentthelife
cycle phase. After the modification, the segmentation was simplified into ‘A-fast’,
‘A-slow’, ‘B-fast’, ‘B-slow’ and so on to roughly reflect the function groups and life
cycle phases.
3.5.2 Integrate and Link Data
After the previous phases of understanding, we realized there was the need to build
two prediction models for the two phases, since the deviation could happen in each
phase and the detection of deviation is necessary to take actions in each phase. For
themodellingofsupplierleadtimedeviation, , theinformationofpartsandsuppliers
were integrated into the records of delivery precision performance Then, for building
the models of transportation lead time, to consider the previous delivery precision
performance from material suppliers could also be beneficial. However, the data of
two phases in the company are independent. They are separated into two systems,
managed by different departments and not linked with each other. In consequence,
there is no information about which parts are carried in the shipments from the
transportation booking. We manually linked the instances from these two phases,
using event time (Dispatch week in material suppliers records, Prove of collection
date in LSP records) and companies (supplier ID in material suppliers records, con-
signor ID in LSP records) as linking keys. When these two keys were in line with
each other in two instances, these two instances were integrated and regarded as the
same ordered flow as Figure 3.3 illustrate. This linkage can help the prediction of
TLT to have more potential features including relevant parts and material suppliers
information.
Another issue is that one transportation booking could contain several ordered
parts, therefore, when left joining parts information into the transportation book-
ing records, several transportation booking instances were duplicated with the only
difference of part information between them. Then, in order to integrate these du-
plicated instances into one independent instance, the information for those parts in
the same transportation booking was used their average value in this project.
3.5.3 Delimitation in the Data Preparation
For data preparation in modelling supplier lead time deviation, normally there are
existing several orders for a spare part with one supplier in two years duration.
Even though the differences between these orders and further integrated features
could be only the event time, the deviation could differ from one order to another
order. Therefore, all the orders kept for input instance for the benefits of repre-
senting the real case, although this might sacrifice the variance of each feature in
eachinstanceandaffectthemodelperformanceandtheresultoffeatureimportance.
25 |
Chalmers University of Technology | 3. Methods
Supplier Lead Time
Transportation
Deviation
Delivery Precision
Performance
# Supplier ID # Consignor ID
# Dispatch week # Prove of collection date
Integrated Inbound
Delivery Flow
Figure 3.3: Integrate and link data for the phases of material supply and trans-
portation
There are three ways of transportation, namely Door to Door (DDT), Cross-docking
and Milk run. For the transportation mode using cross-docking, the transporta-
tion booking reservation is separated into two independent transportation booking
records. The previous cross-docking is from material supplier to cross-dock point,
while the later cross-docking process starts from cross-docking point to CDC. The
consignor for the second transportation booking records, therefore, becomes the
cross-dock point. In this way, the second phase of cross-docking transportation
failed to be linked with previous corresponding records of material suppliers due to
the key of supplier ID and consignor was not to be matched. Only the previous leg
of cross-docking were linked.
Another limitation happened for the milk run transportation. Even though one milk
run generates one transportation booking, with the two keys can be in line with the
first material supplier in the milk run, the information of the remaining suppliers
and parts information failed to be considered into the input instance for the milk
run transportation. As Figure 3.4 shows.
3.5.4 Feature Selection
To represent previous identified factors into candidate features for modelling, there
were a few cases occurred in this process. Firstly, there are data which can directly
represent the factors such as the demand, value, stackable, hazardous, custom, eval-
uation results for material suppliers. Secondly, there were data representing the fac-
tors at an aggregated level, such as TB weight and volume data for the total weight
and volume in one shipment, segmentation data for integrating function groups and
life cycles, country for traffic and weather. Thirdly, some factors that were not
recorded in the data form, such as the prioritization. Some factors’ information is
26 |
Chalmers University of Technology | 3. Methods
not available in the buying company due to that information is owned by material
sup- pliers such as material suppliers’ production information. These factors were
tried to be indirectly reflected by other available data, such as sales spend level data
on suppliers for representing the prioritization, quality and environment certificate
for representing the production capacity of suppliers. However, some data currently
are not integrated into the database, and we could not either find other suitable
data for the indirect representation of their corresponding factors, such as historical
delivery precision performance and evaluation results of LSP.
Sincethedimensionofinputfeaturesinthisprojectislimited,allpotentialcandidate
features were kept as input for the modelling. No further feature selection is needed
for the benefits of dimension reduction which is not the case with limited feature
dimension.
Material Supplier 1
CDC
Material Supplier 2
Material Supplier 3
(A)
Material Supplier 1 CDC
Material Supplier 2
Cross-docking points
Material Supplier 3
(B)
CDC
Material Supplier 1 Material Supplier 2 Material Supplier 3
(C)
Figure 3.4: The data linkage in different transportation modes (A) Door to door;
(B) Cross dock; (C) Milk run
27 |
Chalmers University of Technology | 3. Methods
3.5.5 Handling Missing Data
Missing data imputation is a method for filling the missing values with some prob-
able and possible values before the process of learning algorithm begins (Lepping,
2018). Replacing each missing value for a variable by using the average observed
values for that variable is a common method that may accurately predict the value
of the missing data but, also leads to poor estimation of variances and correlations
(Schafer and Graham, 2002). There was a proportion of missing value when we
examined the extracting result. For supplier phase, these missing data particularly
exist in the evaluation information for suppliers, including the Supplier Evaluation
Measurement (SEM) result, logistics audit result and historical delivery precision.
There could be several reasons for the missing value. For example, no evaluation has
been performed or no more cooperation with those material suppliers. The degree
of missing data for supplier phase was presented in Table 3.3. In comparison, for
the transportation model in the data preparation stage, only successful linked and
integrated records were kept, and therefore there is no missing value. The missing
data were filled in with mean value in this project.
Table 3.3: Missing value for supplier lead time phase
Variables Number of instances Missing rate (%)
Dispatch Week 400641 0.00
Part No 400641 0.00
Supplier No 400641 0.00
Lead time deviation 400641 0.00
Parameter reference 388011 3.15
SEM result 288761 27.93
QPM score 399906 0.18
Quality Certificate 329809 17.68
Purchase agreement 400641 0.00
Sales level Spend 399906 0.18
Vital 400641 0.00
Hazardous Code 400641 0.00
Prepacking Type 400641 0.00
Country 400154 0.12
Registration Date 398617 0.51
Stand Price 398617 0.51
Order Hits Roll 13 Period 398617 0.51
Delivery Precision 362068 9.63
Logistics Audit Result 262179 34.56
28 |
Chalmers University of Technology | 3. Methods
So far, a list of features has been constructed as input data for modelling and re-
search question two were covered.
3.6 Machine Learning Modelling and Evaluation
Different machine learning modelling and techniques can be chosen and tested in
the modelling phase. The parameters are required tuned into the optimal values.
Noted the modelling is also closely linked to its previous phase of data preparation
since the new problems of data set could not be unveiled until modelling or new
ideas are generated for collecting new data.
Thefirstchoiceinthemodellingistochoosefromsupervisedlearning,semi-supervised
learning and unsupervised learning (Lingitz et al., 2018). Since the purpose of the
project work is to predict the lead time deviation as the output with labelled input
data from databases, supervised machine learning was used for this situation.
Based on the previous understanding of the business goal and data mining goal
(Smith-Miles, 2009), the output variable is made into three classes, namely ‘On
time’, ‘Early’, ‘Late’. This is an imbalanced data set with the majority of the
observation falling into the ‘On time’ class. Balanced Random forest (Chen and
Breiman, 2004) and boosting algorithms (Kotsiantis, Kanellopoulos and Pintelas,
2006) could be two approaches to deal with imbalanced data set. In addition, based
on the knowledge from the data scientist in the case company, several classifica-
tion machine learning algorithms were selected to build the models for each phase,
including Balanced random forest, Catboost and Gradient boosting. Balanced ran-
dom forest has been selected as the algorithm is combining the bagging method and
under-samplingtechniqueforthemajorityclass(ChenandBreiman, 2004). Therea-
son for selecting the Catboost and Gradient Boosting is that both of them are using
the boosting method which can give high penalty to missing classified minority class
asacost-sensitivelearningtechnique(Kotsiantis, KanellopoulosandPintelas, 2006).
Finally, an evaluation process was conducted. The performances of the above- con-
structed models were compared and recorded using confusion matrix. The results
were analyzed from a data analysis point of view. Furthermore, the improvement
anddeploymentofthemodelswereexaminedconsideringthefulfillmentthebusiness
goal. The process of the CRISP-DM model was reviewed. Future possible actions
were proposed. Until this point, the research question 4 was answered. The rela-
tionshipbetweenprocessesandresearchquestionsareillustratedbelowinFigure3.5.
29 |
Chalmers University of Technology | 3. Methods
RQ 1: What are the benefits of predicting lead
time deviation for buying companies?
Business Understanding
RQ 2: What are the factors that could be
associated with lead time deviation perceived
by buying companies?
RQ3: Which data are available to be used as
Data Understanding
features when building the prediction model
& Preparation
of lead time deviation at Volvo SML?
RQ4: How should the prediction model be built
Modelling using machine learning considering the practicality
of use in the current stage at Volvo SML?
Figure 3.5: The relationship between processes and research questions
3.7 Validity and Reliability
According to Bryman and Bell (2011), there are two important aspects regarding
the evaluation of the quality of research, namely reliability and validity. Reliability
is about the consistency of measures, whereas validity refers to whether a measure
of a concept actually manages to measure it (Bryman and Bell, 2011).
In the qualitative part of this thesis, reliability will be increased by contemplating
inter-observer consistency. According to Bryman and Bell (2011), inter-observer
consistency is an issue of inconsistent declaration that could happen when there are
several observer-constellations judging information subjectively. All the interpreta-
tion from interviews were analysed and agreed upon by the presented interviewers.
Validity in the qualitative data of research would increase through internal validity,
it means that the findings from observations should fit into the theoretical frame-
work developed (Bryman and Bell, 2011). This subject was considered during the
thesis process in order to verify the findings from interviews with actual modelling
further on.
During the quantitative data of the thesis, face and convergent validity were con-
sidered. According to Bryman and Bell (2011), face validity is about the process
of evaluation of a model by an outside expert to see if it is reasonable. Based on
this factor, a machine learning expert from the department where the thesis project
is conducting evaluated the scientific aspect of machine learning algorithms in the
30 |
Chalmers University of Technology | 4
Results: Business Understanding
The first section in this chapter is going to describe the company’s operation around
lead time including involved processes and roles. Then, the current performance of
lead time deviation and its impact are presented. Furthermore, the business goal for
this data mining project is set. Finally, the factors related to lead time deviation are
described.
4.1 The Set-up of Lead Time in Volvo
At Volvo SML, most of the lead times are negotiated with material suppliers and
LSP. As agreed, these lead times will be set as predefined parameters in the planning
systems. The supply process in Volvo SML could be categorized into five processes,
asFigure4.1shows. InboundsupplyphasestartsfromContinentalMaterialPlanner
(CMP) placing orders to material suppliers and ends till the orders are received and
registered at Central Distribution Center (CDC), including supplier lead time, in-
bound transportation lead time and internal receiving lead time. Outbound supply
phase begins after CDC have received and registered the orders until customers get
their requested spare parts including outbound transportation lead time and order
lead time. The shipments are carried by LSP.
Since the set up lead time between Volvo and suppliers by negotiation is an esti-
mation of lead time, together with other causes of disruption alongside the delivery
process, the deviation in lead time is inevitable. There are also cascading effects
along the supply chain. For example, when the material supplier does not dispatch
the orders on agreed time, that is going to affect LSP on picking up the orders and
further affect later process of transportation. The affected trucks may further arrive
at CDC later than the schedule and may need to wait to be unloaded since the
capacity of CDC is limited. Most importantly, currently there is no existing process
or tool to predict the deviation of lead time in the company.
33 |
Chalmers University of Technology | 4. Results: Business Understanding
Supplier
A Dealer A
RDC
Supplier LSP1
Dealer B
B
CDC
Supplier LSP2 Dealer C
C
Sup Dplier LSP3 SDC Dealer D
Dealer X
Supplier
X
Inbound Outbound
\
Supplier Inbound Internal Outbound
lead transportation receiving transportation Order
time lead time lead lead lead
time time time
Inbound flow
lead time
Figure 4.1: The set up of lead time in Volvo SML
4.2 The Process and Roles Involved in Dealing
with Lead Time Deviation
The process and roles involved in dealing with lead time prediction are introduced
in the below sections. These results lead to the setting of business goal.
4.2.1 Process Overview
The inbound delivery process behind SLT starts from Demand and Inventory Plan-
ners(DIP) generate demandforecastfor CDCGhent. The demandforecastcontains
information about at what time and how much of which spare part is needed in the
CDC. These demand forecasts pass through the planning system. Based on the
forecast information, CMP place orders to corresponding material suppliers. When
material suppliers are ready to dispatch the order, they book the shipments from
LSP through Volvo’s transportation management portal ‘Atlas’. The transportation
booking (TB) contains information such as pick up and shipping address, volume,
weight of spare parts. LSP will ship the order to the CDC based on transporta-
tion booking information scheduled by Atlas. Atlas portal also incorporates the
transportation orders from several material suppliers by arranging different ways of
delivery including DDT, cross docking and milk run.
34 |
Chalmers University of Technology | 4. Results: Business Understanding
Supplier Relationship Managers
Lead Time Follow Up Escalate
Suppliers Continental Material Planners
Supplier Place Order
lead
time
Lead Time Monitor and Review
Planning Demand
Atlas Specialist
Transportation Booking Demand and Inventory Planners
Transportation Booking
Inbound Logistics Service Providers Transport Material Coordinator
logistics Lead Time Monitor
lead
time
Lead Time Review
Lead Time
Optimize
Manager Supplier Management
Transport Developers
Lead Time Follow Up Escalate
Figure 4.2: The roles involved in dealing with lead time deviation
Noted the role description is in line with current responsibilities, which could be
changed from time to time. The following section is going to describe in detail the
responsibility of the most relevant roles, that are the monitors and evaluators of
lead time deviation, including Continental Material Planner (CMP) , Supplier Re-
lationship Manager (SRM), Supplier Manager (SM). For managing material suppli-
ers, Volvo has CMP for monitoring the individual level of performance on material
suppliers and SRM perform a higher integrated level of management. While for
transportation, TMC are responsible for managing the individual level of LSP and
SM are for a higher level of measurement. Delivery precision measures whether the
suppliers dispatch requested order on the scheduled time and this key performance
indicator (KPI) directly links to the degree of deviation on SLT. Similarly, there is
also delivery precision measuring the transportation lead time deviation from LSP
representing the accuracy of ETA. The information about the key roles and KPIs
for lead time performance is summarized in Table 4.1.
Table 4.1: The key roles and KPIs for lead time performance
Suppliers KPI of lead time Key Roles
performance
Material suppliers Delivery precision Continental Material Planner (Monitor)
Supplier Relationship Manager (Evaluator)
Logistic service Delivery precision Transport Material Coordinator (Monitor)
providers Supplier Manager (Evaluator)
35 |
Chalmers University of Technology | 4. Results: Business Understanding
4.2.2 Continental Material Planner
CMP are responsible for the inbound material supply process for spare parts. Their
mission is to ensure the availability of spare parts at the central warehouse and
provide a sharp ETA to the customers. Their first responsibility is to set up SLT
with material suppliers when the part is first sourced to them and then to review
the lead time after a certain period of time. The guideline is to propose 2 weeks
of lead time for high running spare parts which are frequently ordered, 4 weeks for
the middle runner, and best possible lead time for low runners. If proposal for SLT
is not accepted by material suppliers, then CMP will take what material suppliers
answer to them. Lead time review is done once or twice with two material suppliers
per year for each CMP. The purpose of lead time review is to shorten lead time
and have lead time information alignment with suppliers. SLT is important since
it determines the amount of safety stock. Besides, during the period of SLT, CMP
cannot change the order from suppliers unless the change is agreed by suppliers.
Continental Material Planner
Create & Send
Delivery Schedule
Purchase Order
Follow up Supplier
Dispatch
Logistic Preparation
Parameters
Analyze & Decide No Delivery Schedule Supply According No Solve Delivery
Corrective Action Covers Demand? to Plan? Deviation
Yes Yes
Escalation Yes
Needed?
No
Delivery Schedule Closed Delivery Schedule Closed
Figure 4.3: The working procedure of CMP
Another important responsibility of CMP is to place the order to material suppliers
based on purchase orders from DIP and logistic preparation parameter set in the
system. After placing the order, CMP then monitor suppliers’ delivery precision
by having frequent contacts with them. If suppliers confirm the order information,
CMP send the information of ETA to the following process. If there is deviation
36 |
Chalmers University of Technology | 4. Results: Business Understanding
happened in the material suppliers, CMP are responsible to figure out the reasons
for the deviation and take actions for dealing with deviation. For example, if the
order is dispatched later than schedules, CMP can arrange extra transport with the
rush option in order to ensure the availability of spare parts. Since the rush trans-
port causes high costs, only with critical spare parts and backorder from customers,
CMP shall use this option. CMP can also decide to escalate the problematic suppli-
ers to SRM where re-examinations of the suppliers will be performed. In contrast,
if one supplier’s performance is above a certain percentage for a certain period of
time, CMP tend to trust this supplier and may send out the ETA information very
soon without confirmation from material suppliers. The process is illustrated above
in Figure 4.3.
4.2.3 Supplier Relationship Manager
SRM take responsibility for supporting and developing material supplier in the field
of logistics by evaluating supplier delivery performance. SRM are also in charge of
conducting Materials Management Operational Guidelines / Logistics Evaluation
(MMOG/LE) audit. The purpose of this audit is to evaluate the logistics maturity
of material supplier and initiate an action plan for identified gaps. This audit has
three levels namely supplier self-assessment, desk verification of a self-assessment
and on-site verification. Specifically, in the audit, there is a document of evaluating
suppliers performance purely on logistics including lead time agreement, value, ma-
terial handling, organization, production, communication, planning of all logistics.
Material suppliers fill in the report and SRM have a site visit to evaluate these
performances when necessary.
SRM are also managing low performing suppliers, if these suppliers performance are
not improved for an agreed period of time, SRM should escalate them to supplier
purchasingdepartmentandthesematerialsuppliersmayenduplosingcontractfrom
Volvo. Another task of SRM is prioritizing deliveries between Volvo manufacturing
sites and CDC when there is crisis such as lack of capacity in material suppliers.
Critical spare parts are among the first priority, and then the manufacturing sites
get their capacity, finally, the non-critical spare parts get the rest of capacity.
4.2.4 Transport Material Coordinator
Similar to the responsibility of CMP on material suppliers, TMC is responsible for
monitoring the performance of LSP in terms of agreed procedure and targets. For
their appointed distribution flow including DDT and milk run, they are following
up the performance indicators agreed upon with LSP while cross-docking transports
are managed by another specialist.
37 |
Chalmers University of Technology | 4. Results: Business Understanding
If deviations happen, TMC also need to analyze the cause of deviations and take
corrective actions within their responsibility area or propose corrective actions out
of their responsible area. For example, if material suppliers cause the deviation,
they should be escalated by TMC. If the deviation is caused by LSP, TMC could
take corrective plan or escalate them to SM. This process is demonstrated below as
Figure 4.4.
Transport Coordinator
Transport Collection Transport Performed
Monitor & React on Group Analyze & Log
Transport Deviations Material Supplier
Manager Specialist
Deviation Perform Material
Material
cause? Supplier Root Cause
Supplier
Analysis
Multileg
Transport
type?
Cross Dock Support
Specialist
Performance followed up
Direct flow or delivery to Logistics Service
ultimate consignee Provider
Supplier Manager
Monitor & React on Perform Logistics
Arrival Service Provider
Root Cause Analysis
End Performance followed up
Figure 4.4: The working procedure of TMC
4.2.5 Supplier Manager
OneoftheresponsibilitiesthatSMhaveisthequalityassuranceforLSP.Thismeans
that SM have to make sure that every appointed LSP will deliver the agreed level
of delivery performance based on their contract. There are some predefined targets
related to the service levels for the LSP, such as pickup and delivery precision, their
communication performance regarding reporting deviation in time. Following up
these targets, making improvements and reporting them in terms of different weekly
and monthly KPI are SM’s tasks. It means that they follow up the performance of
LSP in terms of delivery precision.
38 |
Chalmers University of Technology | 4. Results: Business Understanding
For those delivery deviations, SM are required to perform root cause analysis and
take correction plan accordingly, in order to avoid or limit the consequence of de-
viation. For example, due to the dynamic character of the business environments,
there would be disruptions such as harbour strike, storms, which would affect the
planning. Efficient crisis management for them is a must to solve the problem in a
short time and be sure that the planning schedule would not be affected too much.
One of the solutions SM are using is to arrange meetings with LSP. The objective
of these arrangements is to analyse the new situation and agree upon the standards
and performance expectations based on new conditions in an open, straightforward
and easily understood way to finally reach the target.
4.3 Situation of Deviation
Figure 4.5 shows the average SLT deviation of all spare parts for Volvo truck during
the period of 2017 and 2018. The negative value represents the length of early
dispatched orders in week (s) while the positive value represents the late ones. As
the figure shows, there is one fluctuation in performance happened at the end of
2017, where large deviation occurred. The reason for this fluctuation is because
this period corresponds to the Christmas break when the material suppliers cease
production and operation. Otherwise, the delivery precision for truck spare parts
has no seas onal trend.
Average Supplier Lead Time Deviation
10
8
6
4
2
0
Total
-2
-4
-6
-8
-10
-12
Figure 4.5: Average SLT deviation deviation for 2017-2018
The goal of delivery precision for material suppliers in Volvo is 95%, that contains
all the dispatches not being late (including early and on time). Figure 4.6 shows
that for the past two years, this actual delivery precision of not being late is 86%.
Besides, among this 86%, up to 9% of the order dispatched earlier than scheduled.
There is a significant gap between the goal and current deviation of both late and
early delivery.
39
307102 807102 317102 817102 327102 827102 337102 837102 347102 847102 108102 608102 118102 618102 128102 628102 138102 638102 148102 648102 |
Chalmers University of Technology | 4. Results: Business Understanding
Delivery Precision of Material Suppliers
9%
14%
Early
Late
On Time
77%
Figure 4.6: Delivery precision of material suppliers for 2017 and 2018
The goal of delivery precision for LSP in Volvo is 97%. However, for the trans-
portation of the spare parts to Ghent CDC for past rolling one year, only 90% of
them was not delivered late as Figure 4.7 shows. Further, 27% out of 90% actually
delivered earlier than expected. The deviation of transportation is even larger than
the previous delivery performance of material suppliers.
Delivery Precision of LSP
27%
EARLY
LATE
ON TIME
63%
10%
Figure 4.7: Delivery precision of LSP for past one year from 2019
4.4 Impacts of Lead Time Deviation
The deviation of lead time could bring various side effects and deteriorate the com-
pany’s performance. These potential effects can be closely examined when the de-
viation occurs in material suppliers and LSP in terms of late and early delivery
respectively.
When the spare parts cannot be dispatched on time according to the schedule from
material suppliers, the immediate consequence could be the waste of transportation
when LSP go to material suppliers based on TB information but end up failing to
pick up the requested order. Even if the material suppliers communicate well about
the delay information and change the new transport booking, the parts still arrive
late at CDC Ghent. This could result in loss of availability when there is a demand
for those parts, which means the company will fail to deliver what is requested due
to lack of inventory. Likewise, the late delivery of LSP directly affects the stock
40 |
Chalmers University of Technology | 4. Results: Business Understanding
level in CDC, and could further impact availability of stock possibly. This conse-
quence could also cascade till the rest of the supply chain including the availability
in regional distribution center (RDC) and dealers. Finally, it impacts customer sat-
isfaction. In order to maintain the availability of spare parts, the cost is to adopt
rush transportation such as air which is bringing in the high cost of transporting
freight. The cost of rush air is huge for Volvo SML.
The spare parts could also be dispatched earlier than ordered from material suppli-
ers. This is because on some occasion when they finish producing the orders earlier
or have the stocks of the requested order. They could choose to book transportation
in Atlas platform and push the parts to Volvo in order to get rid of their stock. In
addition, the deviation in TTL could also result in less transportation time than
estimated. These early arrived parts could also bring some problems. They could
disturb the operation in CDC since the capacity of unloading and storage in a ware-
house is limited. These early arrived parts could be waiting to be scheduled capacity
and then unloaded at CDC. Furthermore, the inventory cost and tied up capital of
keeping the parts for a longer time will also increase.
4.5 Business Goal of the Project
From the investigation, the lead time is a predefined parameter in the planning sys-
tem, and it is set by the negotiation between the suppliers and Volvo. The lead
time is very static in the system which is reviewed and updated relatively infre-
quently. There is a fairly high proportion of deviation existing in the performance
of lead time. There are two roles in Volvo (CMP, TMC) who are directly responsible
for monitoring the performance of suppliers regarding lead time deviation and tak-
ing action accordingly, which is achieved by close communication with each supplier.
However, to proactively communicate with all suppliers is time-consuming and less
effective. To wait for information from suppliers about their deviation situation is
not very reliable which depends on suppliers’ proactiveness. Therefore, if Volvo can
predict the lead time deviation in advance, it could be used as a deviation alert for
the monitors. These monitors could selectively pay more attention to the suppliers
that are predicted to have deviation. Communication can be more effective between
monitors and suppliers to detect the deviation. This could help to trigger the ac-
tions to prevent the happening of deviation in advance. For other cases where the
deviation is confirmed and unpreventable, the monitors can reschedule the inventory
to correct the deviation. To sum up, the business goal for this project is to gener-
ate a deviation alert created by predicting lead time deviation of certain suppliers
for certain orders. This alert could be used by CMP and TMC to be precautious
and proactively contact the suppliers with deviation alert. This could improve the
precision of ETA and ensure the availability at a low cost.
41 |
Chalmers University of Technology | 4. Results: Business Understanding
4.6 Factors Related to Lead Time Deviation and
the Availability of Data
Gathered from the literature study, interviews as well as internal documents, a pile
of factors that could be associated with lead time deviation in Volvo are compiled
below, the availability of their corresponding data is also examined.
4.6.1 Factors of Material Suppliers’ Lead Time Deviation
The sources of deviation can be categorized into two aspects, namely parts and sup-
pliers. The deviation from parts is associated with their characteristics such as the
complexity of producing spare parts, demand on the spare parts, and the criticality
of spare parts. Most of the factors belong to correlation factors which could have
associations with deviation but not directly result in deviation.
Deviation related to parts:
The characteristics of spare parts, including the function groups, the life cycle posi-
tion, the demand for spare parts, criticality of spare parts and value of spare parts.
The function groups refer to the parts of the vehicles where the spare parts belong
to. For example, the engine, fuel system, brake belong to different functions. This
function group could reflect the complexity of the production. The production of
the engine is more complicated than producing brake. The risks of suffering from
deviation in production for the engine could be higher than those for the brake. The
function group could also reflect its components of raw material. Since for automo-
tive manufacturing, material suppliers rely heavily on their suppliers for providing
the raw material. The supply situation of raw material could affect the production.
For example, crisis of metal happens frequently than the plastics. There are 2882
function groups at Volvo Group.
The life cycle position is a changing statue in the life cycle starting from the intro-
duction, to phase out of a certain truck model. It is determined by the number of
years since introduction. The demand for spare parts corresponds to the life cycle
phase of their related model. When more trucks sold, more related spare parts will
be in need for that model and vice versa. For example, when a truck model is going
through phase out, the stock for its spare parts is needed only for serving existing
vehicles. There will be a decreasing demand gradually in the future. The demand
for spare parts directly give the information on the amount of parts that have been
ordered. However, the impacts of demand on deviation are uncertain. The higher
order amount could bring in the economy of scale for production and draw more
attention from production scheduling, therefore, decrease the risk of having devia-
tion, whereas producing a large amount of parts could bring in risks of production
disruption.
42 |
Chalmers University of Technology | 4. Results: Business Understanding
The criticality of spare parts. Different spare parts could have different criticality
regarding their importance to secure up time for the vehicles. The higher criticality
of spare parts derives from their higher importance to make the vehicle functions
well. For example, a broken part which results in engine failure is more critical
than a broken back mirror. Since the produced parts from one material supplier
could be used for production sites, service market and powertrain in Volvo. When
suppliers do not have the capacity for production from all demand, SRM do the
site share based on the priority scale. The spare parts with high criticality are the
first priority, and then the production parts get their capacity, finally the rest of
the spare parts. Therefore, the non-critical parts could be more likely to suffer from
deviations in delivery when there is limited capacity. There is a criterion named
Vital code classifying the criticality of spare parts.
The value of spare parts could have a relationship with deviation. Since the expen-
sive parts tend to have higher criticality and could be more important regarding
their costs, it could receive more attention and enjoy higher priority for production.
However, expensive parts tend to be more complicated regarding manufacture and
have higher risk of production disruption. There is standard price recorded at Volvo
for each part that corresponds to the value of spare parts.
In Volvo, there exists a measurement called segmentation that is classified by all
above kinds of characteristics. There are 59 kinds of segmentation currently exist-
ing at Volvo.
Deviation related to Material suppliers :
There are several factors related to material suppliers. One factor is the production
disruption occurred at the suppliers manufacturing site that directly causes devi-
ation. There are also supplier production capacity, supplier prioritization, supplier
evaluation results and supplier historical delivery performance correlated to the de-
viation of lead time.
Production disruption refers to the disturbances happened in the process of produc-
tion that deviate the production process such as machine breakdown and labour
shortage. In consequence, production disruption results in deviations of lead time.
However, in the company, there is no data available or suitable to represent the
production disruption happened in material suppliers. Meanwhile, supplier produc-
tion capacity, that is referred to the maximum production volume that a supplier
can handle at one time, could reflect the size and furthermore the ability of sup-
pliers to handle production deviation. Suppliers with large capacity tend to deal
with production disturbance smoothly by scheduling resources to bottlenecks and
then dispatch orders properly. There is no direct information available in Volvo for
supplier production capacity. However, it could correlate to sales level spend, order
hits, book off quantity for material suppliers. Since the more money the company
spends on its suppliers and the larger volume the company orders from them, those
suppliers are more likely to be larger firms with a larger capacity. Quality certificates
and environment certificates such as certificate ISO 14000 and QS4000 could also
43 |
Chalmers University of Technology | 4. Results: Business Understanding
reflect different production standards and relate to the ability to handle production
disruption.
Supplier prioritization. A material supplier normally supplies to various customers,
and therefore the production capacity of material suppliers is going to assign to
different buying companies. In case they are overbooked for the production, the
order from the buying company may be delayed due to the suppliers’ prioritization
for other buyers’ order when that company is not their prioritized customers. Volvo
does not have direct information on how prioritized they are as a buyer. Never-
theless, the size of business between Volvo and the material supplier could closely
relate to the prioritization, since suppliers tend to have a closer relationship with
buyers with large orders and set these buyers with a higher priority. As Volvo has
information about sales level spend, order hits, book off quantity, these could be in-
dicationsofthesizeofbusinessandtherebytheclosenessofthebusinessrelationship.
Supplier evaluation result. From a buying company’s perspective, some performance
of suppliers can be perceived and recorded, and this information on suppliers could
be evaluated. The knowledge is obtained in order to evaluate and develop sup-
plier performance and make the decision for further cooperation. Since the eval-
uation information could be closely linked to the suppliers’ delivery performance,
this evaluation information could be used to predict future deviation. In Volvo,
SRM perform supplier evaluation and generate SEM results. This result evaluates
the overall performance of a supplier by examining various aspects including com-
pany profile, management, environment, quality, logistics, aftermarket, competence,
product development, finance, productivity, and sourcing. The SEM results are
consolidating all the performance and generating one score with the scale between
0-100 for each material supplier in a certain period. In addition, SRM perform lo-
gistic audit specifically. This is an audit of evaluating suppliers’ performance purely
on logistics including lead time agreement, value, material handling, organization,
production, communication, planning information of all logistics aspects. The lo-
gistic audit result could also be reflected by another score consolidation the results
from these criteria. These evaluation results could be a good indication of suppliers’
ability to deliver on time.
Supplier historical delivery precision could be very informative in terms of predict-
ing the future performance of a supplier and lots of traditional prediction methods
are purely based on historical information. At Volvo, delivery precision is a key
performance indicator for material suppliers. It is the percentage of the number
of parts dispatched on time divided by the total number of dispatched parts. The
result indicates the percentage of a material supplier fulfilling orders with the right
quality at the right time with the right paperwork attached.
44 |
Chalmers University of Technology | 4. Results: Business Understanding
Table 4.2: Factors related to lead time deviation in Volvo
Phase Sources Factors
Characteristics (Function groups)
Life cycle phase
Parts Demand
Criticality
Factors of Material
Value
Suppliers
Production disruption
Supplier prioritization
Material suppliers
Supplier evaluation results
Supplier historical delivery precision
Weight, Volume
Stackable
Parts Hazards
Custom
Factors of Inbound
Demand
Transportation Lead
Time Value
LSP Evaluation results
LSP historical delivery precision
Material supplier delivery precision
Transport scheduling
Supply chain Warehouse scheduling
Country
Traffic and weather
4.6.2 Factors of Inbound Transportation Lead Time
For the deviation in inbound TLT, the factors can also come from three aspects.
The first one is the parts considering that transportation is sensitive to its carried
freight. It could include the logistics characteristics of the parts. The demand for
parts could affect transportation scheduling and further the risk of delay. These
factors all belong to correlation factors.
Deviation related to parts
Logistics characteristics of parts refer to the characteristics of parts that could in-
fluence the transportation including the weight, volume, stackable, hazards, the
requirement of custom. The weight, volume and the stackable could affect the
scheduling of transportation. For example, non-stackable parts and high weight or
45 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.