An AI-powered F1 damage detection and race strategy platform that analyzes race footage, generates telemetry data, and provides intelligent pit stop recommendations — all visualized in a stunning 3D interactive car model.

Comprehensive Video Analysis with Gemini 2.0 Flash

• Upload race footage → Get complete damage assessment
• AI-generated synthetic telemetry showing crash impact
• Automatic 3D model updates with highlighted damage
• Natural language pit strategy recommendations

• 🤖 AI-Powered Damage Detection: Uses Google Gemini 2.0 Flash to analyze race footage and detect damage with confidence scores
• 🎨 3D Interactive Car Model: Real-time visualization of damaged parts with severity-based highlighting
• 📊 Synthetic Telemetry Generation: Creates realistic crash event data showing pre/during/post-crash performance
• 💡 Intelligent Decision Engine: Computes lap time loss, cumulative impact, and AI-powered pit stop recommendations
• 💰 Budget Analysis: Tracks repair costs against F1 budget cap constraints
• 🧠 Gemini Reasoning: Natural language explanations for strategic recommendations
• 🎥 Video Processing: Multi-frame analysis of uploaded race footage (MP4, MOV, AVI)

• Next.js 15 - React framework
• TypeScript - Type safety
• Tailwind CSS - Styling with custom Williams x Atlassian theme
• Three.js / React Three Fiber - 3D car visualization
• Recharts - Telemetry graphs
• Framer Motion - Animations

• FastAPI - Python web framework
• Google Gemini 2.0 Flash - Vision AI and reasoning
• OpenCV - Video frame extraction
• NumPy - Telemetry data generation
• Uvicorn - ASGI server

• Node.js 20+
• Python 3.9+
• Google Gemini API key (Get one here)

# Install dependencies
npm install

# Create environment file
cp .env.example .env.local

# Add your Gemini API key to .env.local

# Run development server
npm run dev

Visit http://localhost:3000

# Navigate to backend directory
cd backend

# Create virtual environment
python -m venv venv

# Activate virtual environment
# Windows:
venv\Scripts\activate
# macOS/Linux:
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Create environment file
cp ../.env.example .env

# Add your Gemini API key to .env

# Run backend server
python main.py

Backend API will be available at http://localhost:8000

1. Start the backend server (see Backend Setup above)
2. Start the frontend with npm run dev
3. Upload race footage using the "Upload Video" button
4. View damage detection results in real-time
5. Analyze telemetry data in the bottom graph panel
6. Review recommendations in the decision card
7. Interact with 3D model to inspect damaged parts
8. Simulate pit stops to see predicted outcomes

• Primary: #00205B (Williams Navy Blue)
• Secondary: #00A3E0 (Cyan Blue)
• Warning: #FFC300 (Medium damage)
• Severe: #FF5733 (Critical damage)
• Success: #2ECC71 (Fixed/optimized)
• Neutral: #141414 (Background)

Comprehensive video analysis - Returns damage detection, telemetry data, and pit strategy in one call.

Request: multipart/form-data with video file

Response:

{
  "crash_description": "Contact with barrier in Turn 4 resulted in front wing damage",
  "impact_severity": "High",
  "damage_detection": {
    "timestamp": "2024-01-15T10:30:45",
    "car_id": "RB20",
    "detected_parts": [
      {"part": "front wing", "severity": 0.75, "confidence": 0.90},
      {"part": "left front tire", "severity": 0.45, "confidence": 0.85}
    ]
  },
  "telemetry_data": [
    {"lap": 1, "speed": 310.2, "throttle": 94.5, "temperature": 84.3, "damage_event": false},
    {"lap": 12, "speed": 217.1, "throttle": 47.2, "temperature": 93.8, "damage_event": true}
    // ... 50 laps total
  ],
  "decision": {
    "recommendation": "Pit Now",
    "Δt_lap": 1.215,
    "cum_loss": 46.17,
    "pit_loss": 22.5,
    "budget_impact": 0.00164,
    "suggested_fix": "Replace front wing, left front tire",
    "reasoning": "With severe front wing damage causing 1.2s/lap loss..."
  }
}

Legacy endpoint - damage detection only (see comprehensive endpoint above).

Compute pit stop decision based on damage analysis.

Request:

{
  "damaged_parts": [...],
  "current_lap": 15,
  "total_laps": 58
}

Response:

{
  "Δt_lap": 0.263,
  "cum_loss": 7.36,
  "pit_loss": 22.5,
  "budget_impact": 0.086,
  "recommendation": "Stay Out",
  "suggested_fix": "Replace front wing",
  "reasoning": "..."
}

Upload telemetry CSV data for analysis.

The decision engine calculates:

1. Lap Time Loss = baseline_lap_time × damage_coefficient × severity
2. Cumulative Loss = lap_time_loss × remaining_laps
3. Budget Impact = (part_cost + pit_procedure_cost) / season_budget

Recommendations:

• Pit Now: Cumulative loss > pit stop duration AND remaining laps > 5
• Monitor: Lap time loss > 0.2s AND remaining laps > 10
• Stay Out: Otherwise

HackTx25/
├── app/
│   ├── globals.css
│   ├── layout.tsx
│   └── page.tsx
├── components/
│   ├── CarModel3D.tsx
│   ├── CommandBar.tsx
│   ├── DecisionCard.tsx
│   ├── TelemetryGraph.tsx
│   └── VideoPanel.tsx
├── lib/
│   ├── decision-engine.ts
│   └── types.ts
├── backend/
│   ├── main.py
│   └── requirements.txt
├── tailwind.config.ts
├── next.config.ts
└── package.json

The application includes mock data for demonstration:

• Pre-configured damaged parts (front wing, tire)
• Simulated telemetry with damage events
• Decision engine calculations

• QUICKSTART.md - Quick setup guide (5 minutes)
• API_GUIDE.md - Complete API documentation
• CHANGES_SUMMARY.md - What's new in this version
• QUICK_REFERENCE.md - Command cheat sheet

Test the backend API:

cd backend
python test_api.py                    # Basic tests
python test_api.py path/to/video.mp4  # Test with video

• Real-time live stream ingestion
• Multi-car comparison
• Official F1 telemetry API integration
• Historical race analysis
• Weather impact modeling
• Advanced ML models for damage prediction
• Gemini 2.5 Pro integration when available

MIT

Built for HackTX 2025