RockShield AI — Early Landslide Warning System (SIH Command Center)

🔗 Live Production App: https://rockshield-ai.onrender.com

RockShield AI is a full-stack real-time geotechnical monitoring system designed for open-pit mining safety. It ingests live telemetry (ground vibration, wall movement, slope steepness, water pressure, rainfall) and pipes it to an in-process Random Forest Classifier to forecast landslide hazards and trigger automated site sirens.

---

🚀 Key Platform Features

1. Interactive Mine Wall Grid Map

• An interactive 9-sector mine coordinates map grid (SEC_A01 to SEC_C03) tracking active sensors.
• Color-coded boundaries indicate safety levels (Green: Safe, Yellow: Warning, Red: Hazard Alert).
• Live radar-sweep animations overlay the active monitoring zone.

2. AI Landslide Risk Analysis Checklist

• Swaps paragraph logs for an organized checklist showing the safety status of each sensor indicator:
  • Vibration Check: Ground vibration levels.
  • Wall Movement Check: Wall displacement indicators.
  • Slope Steepness Check: Slope steepness angles.
  • Water Pressure Check: Ground water pressure levels.
• Triggers automated evacuation advisories and sirens when hazard limits are exceeded.

3. Sensor Tuning Overrides & Preset Scenarios

• Manual override sliders allow mining engineers to simulate geophone vibration, wall movement tilt, steepness angles, and soil water pressure.
• One-click presets allow instant testing of:
  • Safe Preset: Nominal baseline levels.
  • Warning Preset: Rising strain parameters.
  • Critical Preset: Pre-landslide failure parameters.

4. Landslide Simulation Time-Travel Playback

• A history playback slider rewinds the dashboard to inspect sensor parameters from past landslide recording events (60 mins ago, 40 mins ago, 20 mins ago).

5. Safety Alert Limits Configuration

• Interactive configuration sliders to set warning trigger points for ground vibration speeds and rainfall volume.
• Crossing these thresholds immediately flashes a system alarm widget.

6. Tactile Parameter Inspection Modals

• Engineers can click any metrics card, sector coordinate, or diagnostic feed to open glassmorphic overlay modals displaying hardware battery health, transmission signal ping latency, and localized telemetry parameters.

7. UAV Drone Monitoring

• Live aerial photo feeds showing active slope monitoring zones.
• Drone diagnostics panel tracking GPS target coordinates, UAV flight velocity, battery charge, and thermal risk hotspots on the pit wall.

8. AI Brain Specifications Card

• Displays real-time machine learning info:
  • Prediction Success Rate: Live accuracy percentage.
  • AI Decision Trees: Displays active estimator count (500 trees).
  • Methodology logs: Full details about algorithm inputs, training samples size, and checked variables.

9. Analytics view & Charts

• Sensor Risk Levels: Normalized area chart tracking multi-sensor curves.
• Main Risk Factors (AI Weight): Bar chart mapping which sensors the AI cares about most.
• AI Vote Distribution: Radar chart mapping class voter percentages.
• Stability Trend Over Time: Historical line chart tracking slide probability levels.

---

🛠️ Technology Stack

• Backend: Node.js, Express, Socket.IO (broadcasting live drifts every 4s), and an in-house custom Random Forest Classifier (written in JS, supporting CART, Gini index, bagging, and feature subsampling).
• Frontend: React 18, Vite, Recharts, Socket.IO Client.
• Aesthetics: Light-theme glassmorphism (translucent white panels, backdrop-filter: blur(20px), subtle hover transforms, Inter Tight typography).

---

💻 Local Installation & Setup

Terminal 1 — Backend (Port 4000)

cd backend
npm install
npm start

Terminal 2 — Frontend (Port 5173 / 5174)

cd frontend
npm install
npm run dev

Open your browser at http://localhost:5173 (or http://localhost:5174 if port 5173 is bound).

---

📂 Project Architecture

• backend/src/dataset.js — Synthetic data generator mirroring open-pit landslide logic.
• backend/src/randomForest.js — Custom, CART-based Random Forest model (no black-box dependencies).
• backend/src/server.js — Express API & Socket.IO server pushing live telemetry.
• frontend/src/App.jsx — Core layout shell, state hooks, and custom click modals.
• frontend/src/components/Header.jsx — Sliding underline navbar, Live status, and notifications dropdown.
• frontend/src/components/SensorControls.jsx — Override sliders, preset scenarios, and custom dropdown menus.
• frontend/src/components/SectorMap.jsx — Wall monitoring map coordinates grid.
• frontend/src/components/AlertPanel.jsx — AI Risk Analysis checklist card.
• frontend/src/components/Charts.jsx — Visualizations (Radar, Area, Bar, Line).
• frontend/src/components/ModelInfoPanel.jsx — AI brain specs card.