Abstract & Overview
In the contemporary landscape, characterized by the robust advancements of IoT and AI in automation system control and monitoring, vehicle condition diagnostic systems, crucial for ensuring safe and efficient vehicle operation and management, are fully aligned with this technological trajectory. This paper outlines a proposed hardware system designed to facilitate the collection of vehicle parameters for IoT applications and the development of a vehicle condition recognition model utilizing Artificial Neural Network (ANN). The model incorporates two distinct training methodologies: the weight detection method and the learning rate adjustment method. A comparative analysis of these two methods will serve as the foundation for selecting a model that demonstrates superior quality, high accuracy, and optimal efficiency in terms of both time and training resources.