⚡ Watt Watch: Sentinel Grid Topology & Anomaly Intelligence ⚡

3-Tier Energy Integrity & Distribution Twin

Watt Watch is a next-generation "Digital Twin" and investigative platform designed to monitor national energy infrastructure. It transforms a simple pole-based system into a deep hierarchical energy grid (Powerplant → Transformer → Pole), leveraging TigerGraph's relational engine to detect anomalies and trace supply-chain integrity in real-time.

---

📌 Project Overview

Energy grids are complex, branching networks where a failure or theft at the "edge" (Pole) cascades up to the "core" (Powerplant). Watt Watch maps these relationships with geospatial precision, allowing operators to visualize the entire supply chain and identify suspicious load patterns using a hybrid heuristic-ML engine.

---

🚀 Key Features

• 📡 3-Tier Grid Hierarchy:
  • Powerplant (Purple): The core generation hubs.
  • Transformer (Orange): The distribution nodes.
  • Pole (Blue/Multicolor): The consumer endpoints.
• 🗺️ Interactive Topology Map: GPU-accelerated rendering of 3,000+ interactive nodes with dynamic connection lines.
• 🔍 Hierarchy Path Tracing: Click any Pole to instantly highlight its supply line through a Transformer back to the originating Powerplant.
• 🤖 Tiered Anomaly Detection:
  • Edge (Pole): Ratio-based load analysis (Actual vs. Expected).
  • Distribution (Transformer): Aggregates downstream loads to detect transformer over-capacity.
  • Core (Powerplant): Monitors total draw to flag grid-level critical status.
• 📊 Dashboard Intelligence: Real-time counters for "System Health" based on hierarchical risk propagation.
• ⚡ Sentinel Map Engine: Zoom-based rendering for connection lines ensures fluid 60FPS navigation even with thousands of edges.

---

🏗️ System Architecture (Sentinel Stack)

1. Graph Layer (TigerGraph):
   • Utilizes Powerplant, Transformer, and Pole vertex types.
   • Relational edges: SUPPLIES (Plant → TX) and DISTRIBUTES (TX → Pole).
2. Logic Layer (FastAPI):
   • Asynchronous Engine: Handles high-volume telemetry ingestion.
   • Hierarchy Orchestrator: Traverses the graph to calculate cascading loads.
3. Data Layer (Python + Pandas):
   • ml/generator.py: Generates a geographically accurate synthetic grid for any city (Default: Delhi Hub).
   • ml/bulk_loader.py: High-speed ingestion (Batch Size: 1000) using JWT-based TigerGraph authentication.
4. Presentation Layer (React + Leaflet):
   • Spatial Engine: Canvas-based rendering for high-density geospatial points.
   • Graph State: React hooks manage the active supply-chain path highlighting.

---

📂 Directory Structure

/watt_watch/
├── backend/                       # FASTAPI COMMAND CENTER
│   ├── main.py                    # REST API Entry Point
│   ├── apply_schema.py            # GSQL Schema Change Infrastructure
│   ├── clear_grid.py              # Topology Reset Utility
│   ├── db/
│   │   └── tigergraph.py          # pyTigerGraph 1.7 Compatibility & Auth Layer
│   ├── services/
│   │   └── tg_service.py          # Hierarchical Anomaly & Path Logic
│   ├── ml/                        # DATA PIPELINE
│   │   ├── generator.py           # Scalable Grid Generator (Plants, TX, Poles)
│   │   └── bulk_loader.py         # Batch Ingestion Engine
│   └── data/                      # LOCAL SNAPSHOTS (.csv)
│
├── frontend/                      # REACT CONSOLE
│   ├── src/
│   │   ├── MainConsole.jsx        # Grid State Orchestrator
│   │   ├── components/            
│   │   │   ├── GridView.jsx       # Map Engine (Hierarchy Rendering)
│   │   │   ├── DashboardView.jsx  # Hierarchical Metrics
│   │   │   └── AnalyticsView.jsx  # Trend Analysis
│   │   └── services/api.js        # API Integration

---

🧠 Anomaly Detection Logic

The Sentinel engine applies a Bottom-Up Propagation Strategy:

1. Pole Level: If $Actual Load > 1.5 \times Expected Load$, a "High Anomaly" alert is raised at the edge.
2. Transformer Level: $Current Load = \sum(Connected Poles)$. If $Current Load > Transformer Capacity$, the transformer is flagged as "Anomaly".
3. Powerplant Level: $Total Draw = \sum(Connected Transformers)$. If $Total Draw > Powerplant Capacity$, a "Critical" status is propagated to the dashboard.

---

🛠️ Setup & Execution

1. Backend Preparation

cd backend
python -m venv venv
# Windows
.\venv\Scripts\activate
# Install deps
pip install -r requirements.txt

2. Synchronization (TigerGraph)

# 1. Update Schema (Add Plants & Transformers)
python apply_schema.py

# 2. Generate and Load the Grid Topology
python ml/generator.py
python ml/bulk_loader.py

3. Start Services

• Backend: uvicorn main:app --reload
• Frontend: npm run dev

---

👤 Author & Development

Hridyansh (Sentinel Developer) Built for the Scaler School of Technology (SST) Watt Watch Challenge.