The integration of artificial intelligence (AI) and machine learning (ML) technologies is rapidly transforming the design, operation, and optimization of electric motorcycles. This review analyzes research published between 2015 and 2024, focusing on how ML algorithms enhance performance, energy efficiency, diagnostics, and charging strategies across four key domains: electric motors, energy storage, charging systems, and electronic components. The review highlights state-of-the-art solutions such as torque and range prediction using LSTM/GRU models, predictive maintenance via CNNs and autoencoders, energy flow control in hybrid battery–supercapacitor systems using reinforcement learning, and federated learning for privacy-preserving embedded applications. Comparative insights reveal quantifiable performance gains over traditional methods, while integrated frameworks are proposed for linking ML diagnostics, Vehicle-to-Grid (V2G) functionalities, and renewable energy integration. The paper concludes with targeted recommendations for future research, including lightweight edge-deployable models, Explainable AI for safety-critical applications, and the fusion of intelligent charging with eco-design principles, aiming to enable intelligent, sustainable, and high-performance electric motorcycle systems.
Pawlik et al. (Tue,) studied this question.