Reproduce “Explainable Time Series Prediction of Tyre Energy in Formula One Race Strategy” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jamie Todd, Junqi Jiang, Aaron Russo, Steffen Winkler, Stuart Sale, Joseph McMillan, Antonio Rago
Researchers from Imperial College London and Mercedes-AMG PETRONAS F1 Team developed explainable time series models to forecast tire energies using proprietary F1 telemetry, demonstrating improved accuracy and interpretable insights for race strategy optimization.
Key insights
- XGBoost achieved the highest prediction accuracy for tire energies, outperforming all deep learning models, though at a significantly higher computational cost.
- The Temporal Fusion Transformer (TFT) was the most accurate deep learning model, showing strong generalization, particularly when trained on green-flag race conditions.
- Explainable AI methods successfully provided actionable insights: TFT's internal interpretability and TIME identified steering angle and pit stops as key drivers of tire energy, while CausalImpact quantified the impact of race events like Virtual Safety Car deployments.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.