Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explainable Time Series Prediction of Tyre Energy in Formula One Race Strategy

Jamie Todd, Junqi Jiang, Aaron Russo, Steffen Winkler, Stuart Sale, Joseph McMillan, Antonio Rago

Abstract

Formula One (F1) race strategy takes place in a high-pressure and fast-paced environment where split-second decisions can drastically affect race results. Two of the core decisions of race strategy are when to make pit stops (i.e. replace the cars' tyres) and which tyre compounds (hard, medium or soft, in normal conditions) to select. The optimal pit stop decisions can be determined by estimating the tyre degradation of these compounds, which in turn can be computed from the energy applied to each tyre, i.e. the tyre energy. In this work, we trained deep learning models, using the Mercedes-AMG PETRONAS F1 team's historic race data consisting of telemetry, to forecast tyre energies during races. Additionally, we fitted XGBoost, a decision tree-based machine learning algorithm, to the same dataset and compared the results, with both giving impressive performance. Furthermore, we incorporated two different explainable AI methods, namely feature importance and counterfactual explanations, to gain insights into the reasoning behind the forecasts. Our contributions thus result in an explainable, automated method which could assist F1 teams in optimising their race strategy.

Explainable Time Series Prediction of Tyre Energy in Formula One Race Strategy

Abstract

Formula One (F1) race strategy takes place in a high-pressure and fast-paced environment where split-second decisions can drastically affect race results. Two of the core decisions of race strategy are when to make pit stops (i.e. replace the cars' tyres) and which tyre compounds (hard, medium or soft, in normal conditions) to select. The optimal pit stop decisions can be determined by estimating the tyre degradation of these compounds, which in turn can be computed from the energy applied to each tyre, i.e. the tyre energy. In this work, we trained deep learning models, using the Mercedes-AMG PETRONAS F1 team's historic race data consisting of telemetry, to forecast tyre energies during races. Additionally, we fitted XGBoost, a decision tree-based machine learning algorithm, to the same dataset and compared the results, with both giving impressive performance. Furthermore, we incorporated two different explainable AI methods, namely feature importance and counterfactual explanations, to gain insights into the reasoning behind the forecasts. Our contributions thus result in an explainable, automated method which could assist F1 teams in optimising their race strategy.
Paper Structure (12 sections, 9 figures, 5 tables)

This paper contains 12 sections, 9 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Time Series Forecasting of Tyre Energy
  4. Time Series Data
  5. Track State Encodings
  6. Dataset Partitioning
  7. Problem Formulation
  8. Forecasting Models and XAI Methods
  9. Evaluation and Discussion
  10. Time Series Forecasting
  11. Explainable AI
  12. Conclusions and Future Work

Figures (9)

  • Figure 1: Green-flag XGBoost (left) and TFT (right) front-left tyre energy forecasts for one lap of a seen open track
  • Figure 2: Green-flag XGBoost (left) and TFT (right) front-right tyre energy forecasts for one lap of an unseen street circuit
  • Figure 3: XGBoost (left) and TFT (right) SMAPE values over several laps of the Abu Dhabi Grand Prix (a seen open track)
  • Figure 4: XGBoost (left) and TFT (right) SMAPE values over several laps of the Miami Grand Prix (an unseen street circuit)
  • Figure 5: Green-flag TFT all tyre energy forecasts for one lap of an unseen clockwise open circuit. The top row displays the front left and right tyres, while the bottom row displays the rear left and right tyres.
  • ...and 4 more figures