app.py

# app.py - FactoryGPT: FactoryGPT Humanization for the Factories of the Future in Industry 5.0 – Predict, Perfect, and Connect

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sentence_transformers import SentenceTransformer
from transformers import pipeline
from sklearn.ensemble import IsolationForest
import base64
from io import BytesIO
from fpdf import FPDF

st.set_page_config(page_title="FactoryRAG+ - Smart Sensor Twin", layout="wide")
st.title("🏭 FactoryRAG+: Smart Dashboard with AI Monitoring, PDF Reporting & Digital Twin")

# Load models
EMBED_MODEL = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
GEN_MODEL = pipeline('text2text-generation', model='google/flan-t5-base')

# Upload data
uploaded_file = st.sidebar.file_uploader("📂 Upload your condition monitoring CSV", type=["csv"])

if uploaded_file:
    df = pd.read_csv(uploaded_file)
    numeric_cols = df.select_dtypes(include=np.number).columns.tolist()
    st.success("✅ Data loaded successfully!")

    st.subheader("📊 Sensor Snapshot")
    st.dataframe(df.head())

    # --- Multi-signal Dashboard (Collapsible) ---
    with st.expander("📈 Expand to View Sensor Dashboard", expanded=False):
        selected_cols = st.multiselect("Select signals to visualize", numeric_cols, default=numeric_cols[:3])
        fig, ax = plt.subplots(len(selected_cols), 1, figsize=(8, 2 * len(selected_cols)))
        if len(selected_cols) == 1:
            ax = [ax]
        for i, col in enumerate(selected_cols):
            ax[i].plot(df[col], label=col)
            ax[i].set_ylabel(col)
            ax[i].legend()
        st.pyplot(fig)

    # --- Convert Logs to Chunks ---
    def convert_to_chunks(df):
        chunks = []
        for idx, row in df.iterrows():
            sentence = f"[Log {idx}] " + ", ".join([f"{col}: {row[col]:.2f}" for col in numeric_cols])
            chunks.append(sentence)
        return chunks

    if 'chunks' not in st.session_state or 'embeddings' not in st.session_state:
        chunks = convert_to_chunks(df)
        embeddings = EMBED_MODEL.encode(chunks)
        st.session_state.chunks = chunks
        st.session_state.embeddings = embeddings

    # --- Anomaly Detection ---
    st.subheader("🚨 Anomaly Detection (Isolation Forest)")
    iso = IsolationForest(contamination=0.02)
    anomaly_labels = iso.fit_predict(df[numeric_cols])
    df['anomaly'] = ['❌' if x == -1 else '' for x in anomaly_labels]
    st.dataframe(df[df['anomaly'] == '❌'].head(5))

    # --- Digital Twin Summary ---
    st.subheader("🧪 Digital Twin Summary")
    twin_report = ""
    for col in selected_cols:
        max_v = df[col].max()
        min_v = df[col].min()
        mean_v = df[col].mean()
        twin_report += f"{col}\n→ Max: {max_v:.2f}, Min: {min_v:.2f}, Avg: {mean_v:.2f}\n\n"
    st.code(twin_report)

    # --- PDF Export ---
    st.subheader("📤 Export Digital Twin Report as PDF")
    pdf = FPDF()
    pdf.add_page()
    pdf.set_font("Arial", size=12)
    pdf.multi_cell(0, 10, f"FactoryRAG+ Digital Twin Report\n\nSelected Signals: {', '.join(selected_cols)}\n\n" + twin_report)
    pdf_bytes = pdf.output(dest='S').encode('latin1')
    b64 = base64.b64encode(pdf_bytes).decode()
    href = f'<a href="data:application/octet-stream;base64,{b64}" download="digital_twin_report.pdf">📄 Download PDF Report</a>'
    st.markdown(href, unsafe_allow_html=True)

    # --- Role-based Factory Assistant Chatbot ---
    st.subheader("💬 Factory Assistant Chat")
    roles = {
        "Operator": "You are a machine operator. Provide practical insights and safety warnings.",
        "Maintenance": "You are a maintenance technician. Suggest inspections and likely causes of sensor anomalies.",
        "Engineer": "You are a control systems engineer. Offer analytical interpretations and system-level advice."
    }
    role = st.selectbox("👤 Choose your role: Operator, Maintenance, or Engineer", list(roles.keys()))

    if 'chat_history' not in st.session_state:
        st.session_state.chat_history = []

    user_input = st.text_input("Ask FactoryGPT anything (based on uploaded sensor logs):", key="chat_input")

    if user_input:
        query_vec = EMBED_MODEL.encode([user_input])[0]
        sims = np.dot(st.session_state.embeddings, query_vec)
        top_idxs = np.argsort(sims)[-3:][::-1]
        context = "\n".join([st.session_state.chunks[i] for i in top_idxs])
        system_prompt = roles[role]
        full_prompt = f"{system_prompt}\n\nSensor Context:\n{context}\n\nUser Question: {user_input}"
        reply = GEN_MODEL(full_prompt, max_length=256)[0]['generated_text']

        st.session_state.chat_history.append(("You", user_input))
        st.session_state.chat_history.append((f"{role} - FactoryGPT", reply))

    for speaker, msg in st.session_state.chat_history[-10:]:
        st.markdown(f"**{speaker}:** {msg}")

else:
    st.info("👈 Upload a sensor log CSV file to explore digital twin analysis, chatbot Q&A, waveform charts, anomaly detection, and PDF export.")