Toyota GR Cup Racing Replay & Analysis

Interactive Racing Telemetry Replay & AI Coaching System
Real-time race data visualization with intelligent performance analysis

🏁 Project Overview

This is an interactive web-based racing replay application that visualizes Toyota GR Cup telemetry data with an integrated AI racing coach. The system provides comprehensive race analysis, telemetry visualization, and AI-powered coaching insights for racing performance improvement.

🚀 Live Demo

For a live demo, login at: http://52.6.147.127

username: demo

password: TRDHackathon2025Demo@!!

🚀 Core Features

Interactive Race Replay: Real-time visualization of car movement on track with full telemetry data
AI Racing Coach: Powered by Strands Agents SDK with access to comprehensive racing data
PDF Report Generation: Professional performance analysis reports with optimized formatting
Live Telemetry Display: Real-time speed, RPM, throttle, brake, steering, and G-force data
Lap Analysis: Detailed lap timing, sector analysis, and performance comparisons
GPS Trace Visualization: Precise car positioning with racing line overlay
Auto-Detection: Automatic AWS region and account detection for seamless setup

🗺️ Available Data

Datasets are available here: https://trddev.com/hackathon-2025/

The application includes comprehensive Toyota GR Cup racing data from Barber Motorsports Park:

Race Data: R1 (Race 1) and R2 (Race 2) complete datasets
Telemetry: High-frequency GPS, speed, throttle, brake, steering, G-forces
Lap Timing: Sector splits, best laps, lap comparisons, pit stop analysis
Weather Data: Track conditions, temperature, humidity, wind data
Track Layout: 15-turn circuit with sector boundaries and corner analysis

🚀 Quick Start

Prerequisites

Python 3.13+
uv package manager
Git LFS (for large telemetry files)

Installation

# Clone the repository
git clone https://github.com/your-username/trd-hackathon.git
cd trd-hackathon

# Install dependencies
uv sync

# Start the application
python api_server.py

Access the Application

Start the server: python api_server.py
Open browser: Navigate to http://localhost:8001
Select race: Choose R1 or R2 from the dropdown
Select car: Pick a car from the available vehicles
Explore: Use playback controls to replay races and analyze performance

🏗️ Application Architecture

trd-hackathon/
├── api_server.py              # Flask backend with telemetry API
├── racing_agent.py            # Strands AI agent for racing analysis
├── race_replay/               # Frontend web application
│   ├── index.html            # Main application interface
│   ├── track-map.js          # D3.js visualization & controls
│   └── race_maps/            # Track map images
├── dataset/                  # Racing telemetry data (Git LFS)
├── pyproject.toml           # Dependencies and configuration
└── .gitattributes          # Git LFS configuration for data files

🎯 Key Components

Backend (api_server.py)

Flask REST API with CORS support for browser requests
Chunked data loading for smooth telemetry playback
AWS auto-detection using boto3 for seamless cloud deployment
Caching system for optimized data access
Multi-race support with R1/R2 datasets

AI Racing Coach (racing_agent.py)

Strands Agents SDK integration for intelligent analysis
6 specialized tools for comprehensive racing data access:
- Telemetry Analysis (throttle, brake, steering technique)
- Best Laps Data (competitive performance rankings)
- Race Results Analysis (positions, gaps, finishing data)
- Lap Sector Analysis (sector splits, improvements, pit stops)
- Track Position Analysis (corner-by-corner location awareness)
- Weather Conditions (environmental impact on performance)

Frontend (race_replay/)

Interactive track map with D3.js visualization
Real-time telemetry display with live data updates
Video-style controls (play, pause, step forward/backward)
Lap jumping with dropdown navigation
AI chat interface for performance coaching questions
PDF report generation with jsPDF for downloadable analysis
Responsive design for desktop and mobile

📊 Racing Analytics Features

Telemetry Visualization

GPS-based car positioning on accurate track map
Real-time speed, RPM, gear, throttle, and brake display
G-force visualization for cornering and braking analysis
Steering angle and lap distance tracking

Performance Analysis

Best lap identification and comparison
Sector time analysis with improvement tracking
Gap analysis between drivers and optimal performance
Weather impact assessment on lap times

AI Coaching

Context-aware coaching based on current track position
Technique analysis using high-resolution telemetry data
Comparative performance insights vs. field leaders
Strategic recommendations for improvement
Professional PDF reports with comprehensive lap analysis
- Optimized formatting with minimal whitespace
- Color-coded performance indicators (green for achievements, red for critical findings)
- Detailed sector-by-sector breakdowns
- Comparative analysis tables and data visualization
- Downloadable for offline review and sharing

🛠️ API Endpoints

The backend provides comprehensive REST API endpoints:

Race Management

GET /api/races - Available races
GET /api/races/{race_id}/cars - Cars for specific race

Telemetry Data

GET /api/telemetry/{race_id}/{vehicle_id}/timeline - Timeline metadata
GET /api/telemetry/{race_id}/{vehicle_id}/chunk - Chunked telemetry data
GET /api/telemetry/{race_id}/{vehicle_id}/position - Position at timestamp

Performance Data

GET /api/laps/{race_id}/{vehicle_id} - Lap timing and analysis

AI Assistant

GET /api/ai/regions - AWS region auto-detection
POST /api/ai/test-connection - AI agent connection test
POST /api/ai/analyze - Racing performance analysis

🏁 Usage Examples

Interactive Race Replay

Select a race (R1/R2) and car from the dropdowns
Use the timeline slider to scrub through the race
Watch real-time telemetry data as the car moves around the track
Use play/pause controls for automatic playback

AI Coaching

Ask the AI racing coach questions like:

"What could I have done better on this lap?"
"Why was I slower in sector 2?"
"Compare my braking points to the fastest lap"
"How can I improve my cornering technique?"
"Generate a comprehensive performance report for this lap"

Generate PDF Reports: Click the "Generate Report" button in the AI Racing Coach panel to create a professional, downloadable PDF analysis of your current lap performance.

Performance Analysis

Jump to specific laps using the lap dropdown
Compare current lap times to your best lap
Analyze sector splits and identify improvement areas
Review weather conditions that affected performance

🔧 Development

Dependencies

All dependencies are managed through pyproject.toml:

Flask & Flask-CORS for web API
Pandas & NumPy for data processing
Boto3 for AWS integration
Strands Agents for AI coaching
D3.js (CDN) for frontend visualization
jsPDF (CDN) for PDF report generation

Data Management

Large telemetry CSV files are managed using Git LFS:

Configured in .gitattributes for automatic handling
Local data stored in dataset/data_files/barber/
Efficient chunked loading prevents memory issues

📈 Performance Optimizations

Chunked data loading (60-second segments) for smooth playback
In-memory caching with LRU cache for frequently accessed data
Background preloading of upcoming data chunks
Efficient GPS coordinate conversion for track positioning
Optimized telemetry sampling for AI analysis

🌐 Deployment

The application automatically detects AWS environment settings:

Region detection via boto3 Session, EC2 metadata, or environment variables
Account identification using AWS STS for proper context
Fallback mechanisms ensure functionality in any environment

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🤝 Contributing

This racing replay application demonstrates advanced telemetry visualization and AI-powered racing analysis. The codebase showcases integration of modern web technologies with intelligent data processing for motorsports applications.

📞 Contact

TRD Hackathon Team
Project: Interactive Racing Telemetry Replay & AI Coaching System

Built with ❤️ for the racing community

kpx-dev/trd-hackathon

Toyota GR Cup Racing Replay & Analysis

🏁 Project Overview

🚀 Live Demo

🚀 Core Features

🗺️ Available Data

🚀 Quick Start

Prerequisites

Installation

Access the Application

🏗️ Application Architecture

🎯 Key Components

Backend (api_server.py)

AI Racing Coach (racing_agent.py)

Frontend (race_replay/)

📊 Racing Analytics Features

Telemetry Visualization

Performance Analysis

AI Coaching

🛠️ API Endpoints

Race Management

Telemetry Data

Performance Data

AI Assistant

🏁 Usage Examples

Interactive Race Replay

AI Coaching

Performance Analysis

🔧 Development

Dependencies

Data Management

📈 Performance Optimizations

🌐 Deployment

📄 License

🤝 Contributing

📞 Contact

On this page

Languages

Contributors