Abstract
This paper delves into a study of AI-based predictive maintenance in electric vehicles, which has garnered significant interest from research and industrial circles. The advent of AI-based predictive maintenance has a high potential for enhancing the reliability and performance of electric vehicles. It may be employed to predict performance and plan maintenance activities, greatly reducing downtime and maintenance costs. This work conducted research on vehicles where a van-type electric vehicle was monitored and surveyed based on collected data from the cars’ sensors and failures obtained from a diagnostic fault scanner. Data was then collected from ten electric vehicles. After detecting and analyzing car failures, a predictive maintenance prognostic system was created, which may forecast the time to the next breakdown and after breakdowns occur. The AI-based predictive maintenance prognostic system was created using the Weibull regression model, specifically the accelerated life model, to predict the time to come from the Weibull scale parameter. The data were validated and analyzed again to determine the maintenance strategy.
This research also addresses some of the key methods and technologies of AI-based predictive maintenance in electric vehicles, arguing for the importance of predictive maintenance in electric vehicles to avoid early damage, thus contributing to reducing maintenance and repair costs as well as the cost of vehicle ownership. The company will also avoid downtime. At the end of the paper, maintenance strategies for electric vehicles will be demonstrated. The results show the creation of a prognostic model that can predict the time to come, or in other words, the remaining useful life, comparing the actual age of breakdown time. By this time, a company can detect that a part is running out of life in advance to reduce downtime.