AutoGenX: A Hybrid AI-System for Smarter Vehicle Performance and Safety

Modern vehicles are equipped with multiple Electronic Control Units (ECUs), advanced sensors, and interconnected systems that make fault detection a complex task. Conventional diagnostic methods often depend on manual interpretation of On-Board Diagnostics (OBD-II) codes and mechanic experience, which can result in inconsistent outcomes and longer repair times. To overcome these issues, this work presents AutoGenX, a hybrid framework that combines machine learning and generative artificial intelligence for intelligent vehicle fault diagnosis. The system integrates OBD-II data, sensor readings, and user-reported symptoms to identify possible faults and provide simplified repair guidance. Machine learning models such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Transformer-based architectures are used for detection and prediction, while Large Language Models (LLMs) enhanced with Retrieval-Augmented Generation (RAG) are applied to generate user-friendly explanations and repair workflows. Initial testing achieved more than 85% accuracy in fault identification and demonstrated reduced diagnosis time. The proposed framework has the potential to reduce dependency on manual expertise, improve predictive maintenance, and make vehicle servicing more accessible for both owners and mechanics.