Inspiration: What inspired me to make this project was my own experience driving in F1 23. I still have trouble finding the perfect setup and tuning it to something I like more. A lot of drivers give a baseline setup that is very understeery, when I am an aggressive oversteer kind of racer.
What it does: GridTune is an AI-powered F1 setup assistant that generates personalized car setups for F1 25. Users can choose the car, track, weather, session type, and driving preferences, and the app outputs an optimized setup with explanations and a track map.
How we built it: We built GT using Python and Streamlit for the front-end interface. The logic uses rule-based calculations based on track downforce, driving balance, aggression, weather, and session type to generate realistic setup values:
Challenges we ran into: One challenge was making the setup logic realistic while keeping it simple enough for a hackathon. We also had to address issues with image paths, tooltips, and dynamically displaying track maps based on the selected circuit.
Accomplishments that we're proud of: We built a fully interactive, visually polished app in a short time. We also successfully included dynamic setup integration, hover explanations, and automatic track map loading.
What we learned: We learned how to use Streamlit to build an interactive UI quickly and how to organize rule-based engineering logic in Python. We also learned about how F1 car setup changes affect vehicle behavior.
What's next for GridTune: Next, we want to improve setup logic with more advanced parameters, such as suspension geometry and tire wear simulation. We also want to expand to support GT3 setups and telemetry-based recommendations.
Log in or sign up for Devpost to join the conversation.