ABSTRACT
In recent years, deep neural networks have achieved success in Electric Vehicles (EVs) monitoring, primarily due to their scalability with large-scale data and numerous model parameters. However, EVs rely on resource-constrained edge devices that struggle with complex models, and data privacy concerns prevent sharing data outside the owning device. Federated Learning (FL) and Knowledge Distillation (KD) have emerged as key solutions, enabling model simplification and distributed training on private data. FL allows models to be trained locally on edge devices, addressing privacy concerns while keeping data decentralized and avoiding central server dependencies. This approach requires lightweight models optimized for edge intelligence deployment. To address this challenge, we propose architectural solutions leveraging FL, KD and model compression techniques to create simplified Artificial Neural Networks (ANNs) suitable for edge devices in EVs. The proposed architecture integrates these methods into a federated environment, ensuring distributed training while maintaining computational efficiency for EV monitoring and predictive maintenance applications. By combining FL, KD, and model compression, our approach enables efficient and privacypreserving Machine Learning (ML) models, enhancing Edge Intelligence (EI) for EV monitoring in resource-constrained settings.
