app.py

# app.py - FactoryRAG+: AI Assistant with Real-Time Functional Status by Role

import streamlit as st
import pandas as pd
import numpy as np
from sentence_transformers import SentenceTransformer
from transformers import pipeline
from sklearn.ensemble import IsolationForest

# Page config
st.set_page_config(page_title="FactoryRAG+ Assistant", page_icon="🧠", layout="wide")

# Custom dark theme styling
st.markdown("""
    <style>
    html, body, [class*="css"]  {
        font-family: 'Segoe UI', sans-serif;
        background-color: #0f1117;
        color: #f0f0f0;
    }
    .stTextInput>div>div>input,
    .stSelectbox>div>div>div>div {
        background-color: #1a1c23;
        color: #fff;
    }
    .stDataFrame .blank {
        background-color: #0f1117 !important;
    }
    </style>
""", unsafe_allow_html=True)

# Header
st.markdown("""
    <div style='text-align: center;'>
        <h1 style='color: #58a6ff;'>🏭 FactoryRAG+ Assistant</h1>
        <p style='color: #bbb;'>AI-Powered Digital Twin | Real-Time Sensor Health</p>
        <hr style='border-top: 2px solid #888;'>
    </div>
""", unsafe_allow_html=True)

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

# File upload
uploaded_file = st.sidebar.file_uploader("📂 Upload your sensor CSV", type=["csv"])

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

    st.markdown("### 🧾 Sensor Log Preview")
    st.dataframe(df.head(), use_container_width=True)

    def convert_to_chunks(df):
        return [f"[Log {i}] " + ", ".join([f"{col}: {row[col]:.2f}" for col in numeric_cols]) for i, row in df.iterrows()]

    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

    # Condition status based on Isolation Forest
    st.markdown("### ⚙️ Equipment Condition Status")
    iso = IsolationForest(contamination=0.02)
    labels = iso.fit_predict(df[numeric_cols])
    df['status'] = ['❌ No Function' if x == -1 else '✅ Functional' for x in labels]
    df['maintenance'] = ['🔧 Needs Maintenance' if x == -1 else '🟢 Stable' for x in labels]
    st.dataframe(df[['status', 'maintenance'] + numeric_cols].head(), use_container_width=True)

    # Role-based chatbot
    st.markdown("### 💬 Real-Time Role-Based Chat Assistant")
    roles = {
        "Operator": "You are a machine operator. Check if equipment is running properly. If not, flag it immediately.",
        "Maintenance": "You are a maintenance technician. Assess faulty logs and provide service insights.",
        "Engineer": "You are a systems engineer. Offer data-backed advice and failure diagnostics."
    }

    role = st.selectbox("👷 Choose your role", list(roles.keys()))

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

    user_input = st.text_input("🗨️ Ask FactoryGPT about machine status or maintenance needs")

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

    for speaker, msg in st.session_state.chat_history[-10:]:
        st.markdown(f"<div style='margin-bottom: 10px;'><b>{speaker}:</b> {msg}</div>", unsafe_allow_html=True)

else:
    st.info("👈 Upload a CSV file with sensor logs to begin.")