Racing Motorcycle Data Analysis Toolbox for Motorsport Applications

This thesis presents the design and implementation of a modular and scalable data analysis infrastructure for a fully electric racing motorcycle developed by UniBo Motorsport within the MotoStudent competition. The project originates from the need to enhance post-session data interpretation by engineering a flexible and autonomous toolbox capable of generating and visualizing Key Performance Indicators (KPIs) related to vehicle dynamics and powertrain performance. The main objective is to enable data-driven decision-making during testing and developement phases. The methodology combines quantitative data processing with model-based and signal-processing techniques. The system architecture integrates MATLAB for computation and state observers design, MoTeC for trackside visualization, and Power BI for dynamic dashboards. The data workflow spans from raw acquisition to final visualization, including signal conditioning, sensor status verification, and segmentation of telemetry into laps, corners, and dynamic riding phases. Specific algorithms have been developed, such as an Extended Kalman Filter (EKF) for roll and pitch estimation, and a kinematic observer for real-time evaluation of the vehicle geometrical parameters. Several custom KPIs have been defined and validated, including indexes for braking reliability, suspension behavior, geometrical parameters evolution, motor demagnetization, thermal performance, and battery cell unbalance. These indicators have been tested across different circuits and sessions, confirming their robustness and practical utility in assessing the prototype's safety, reliability and performance. The proposed infrastructure has demonstrated strong potential in supporting operational analysis and briefings. Its modular nature allows for future extensions, including model-based estimators and integration with simulation environments, making it a foundational tool for the continuous development of the future racing prototypes.