Inspiration
Watching this F1 season’s rain-affected races drove home how a single pit stop can decide it all. The drama of teams scrambling for the perfect moment to pit—sometimes making or breaking a race—sparked the idea for PitSynapse: a smarter, data-driven way to make these critical decisions.
What it does
PitSynapse is an AI engine for Formula 1 teams. It continuously analyzes telemetry, tire wear, weather, and driver data, simulating thousands of race scenarios per second. Through reinforcement learning, PitSynapse delivers instant, adaptable pit timing and tire choice recommendations—even in unpredictable conditions.
How we built it
- AI/ML: PyTorch & TensorFlow for deep reinforcement learning agents.
- Data: Pandas & NumPy for processing telemetry and weather feeds.
- API Layer: FastAPI for modular, real-time integration.
- Visualization: Plotly dashboards for actionable strategy views.
- Simulation: SimPy and Unity for flexible, circuit-specific scenario generation (future roadmap).
Challenges we ran into
Synchronizing real-time telemetry and weather, modeling rewards for complex strategies, and creating a dashboard that distills actionable recommendations from multi-source data.
Accomplishments that we're proud of
Developing an RL model capable of race-winning pit decisions, building a usable real-time interface, and designing a system robust to data from any circuit or set of race conditions.
What we learned
How tire and weather data truly drive F1 outcomes, practical reinforcement learning for strategy optimization, and the importance of translating complex analytics into tools that real teams can trust.
What's next for PitSynapse
Adding advanced weather prediction, simulating multi-team strategies, expanding scenario diversity, and refining the dashboard for faster in-pitlane decisions.
Log in or sign up for Devpost to join the conversation.