CarMotionData / README.md
danilobarbalace's picture
Create README.md
34a7e33 verified
metadata
{}

Dataset Card for Vehicle Dynamics Dataset

This dataset contains time-series data capturing vehicle motion events such as braking, hard braking, acceleration, and hard acceleration, derived from simulated accelerometer data. It is designed to support research in vehicle dynamics, motion analysis, and driving behavior.

Dataset Details

Dataset Description

The dataset provides accelerometer readings on three axes (X, Y, Z) along with annotated events based on predefined thresholds for braking, hard braking, acceleration, and hard acceleration. It includes data grouped into 10-second intervals for easy aggregation and analysis.

Dataset Sources [optional]

  • Repository: [More Information Needed]
  • Paper [optional]: [More Information Needed]
  • Demo [optional]: [More Information Needed]

Uses

Direct Use

The dataset is suitable for:

  • Training and evaluating machine learning models for detecting vehicle motion events.
  • Analyzing driving behavior under different conditions.
  • Research on vehicle safety systems.

Out-of-Scope Use

  • Real-world deployment without testing with real-world data.
  • Applications outside the context of vehicle dynamics.

Dataset Structure

The dataset contains the following columns:

  • Timestamp: The timestamp of the reading.
  • AccelerationX, AccelerationY, AccelerationZ: Accelerometer readings along the three axes.
  • Braking, HardBraking, Acceleration, HardAcceleration: Boolean flags for motion events.
  • TimeGroup: 10-second interval grouping.
  • Output: The dominant event type in each interval.

Dataset Creation

Curation Rationale

The dataset was created to simulate and analyze vehicle motion events to support research and development in vehicle safety and dynamics.

Source Data

Data Collection and Processing

The data was simulated using randomized accelerometer readings and thresholds for braking and acceleration events. It was then processed and annotated using Python libraries such as Pandas and NumPy.

Annotation process

Annotations were programmatically added based on predefined thresholds for motion events.

Who are the annotators?

The annotations were generated automatically using Python scripts.

Personal and Sensitive Information

The dataset does not contain personal or sensitive information.

Bias, Risks, and Limitations

The dataset is simulated and may not fully represent real-world scenarios. Use caution when applying models trained on this dataset to real-world data.

Recommendations

Users should validate models with real-world data before deploying them in production systems.

Glossary [optional]

  • Braking: A reduction in velocity, identified by negative acceleration.
  • Hard Braking: Significant reduction in velocity, crossing a higher negative threshold.
  • Acceleration: An increase in velocity, identified by positive acceleration.
  • Hard Acceleration: Significant increase in velocity, crossing a higher positive threshold.

More Information [optional]

[More Information Needed]