SIC – Stretto in Carena: An Example of University Innovation in Motorsport

This study presents a detailed analysis of Deep Reinforcement Learning (DRL) for optimizing motorcycle configuration. By integrating advanced machine learning algorithms with complex simulations, DRL provides a powerful tool for identifying optimal configurations that maximize track performance. The training process of DRL models is examined, including the definition of states and actions, the design of reward functions, and the implementation of deep learning algorithms. Furthermore, specific challenges and practical considerations related to the integration of DRL into the decision-making process of teams are discussed, including computation times and the management of data complexity.

PARTECIPANTI: Dario Milone(Faculty Advisor SIC), Massimiliano Chillemi (Vice-Faculty Advisor SIC), Laura Arruzzoli (team leader SIC), Team SIC-Stretto In Carena.