Smart Fault Detection in Electric Vehicles Using Battery and Motor Operation Data Driven Deep Learning

Abstract

As electric vehicles (EVs) become more widespread, early detection of component faults is essential for safety and reliability. This study proposes a smart fault detection system that analyzes daily EV operation data using deep learning models and notifies users of potential defects via a mobile application. We developed two models: an LSTM-based battery fault detection model and a CNN-LSTM-based motor fault classification model. Both models showed strong performance on real-world datasets. For demonstration, the motor model was deployed in a physical setup, achieving near-perfect classification under simulated fault conditions. The results validate the feasibility and potential of deep learning-based fault diagnostics for EVs, emphasizing the need for further real-world validation. © 2025 IEEE.