This paper is a study to improve a system for prediction vehicle failures. To overcome the limitations of the existing deep learning-based vehicle data-based fault prediction system and achieve higher performance, we aim to implement a system that uti...
This paper is a study to improve a system for prediction vehicle failures. To overcome the limitations of the existing deep learning-based vehicle data-based fault prediction system and achieve higher performance, we aim to implement a system that utilizes driving history data and fault diagnosis data to predict faults and notify users. The LSTM-Autoencoder model was used to implement the model, and as a result, the present model showed 94.4% accuracy and 98.6% precision. Through these results, this study achieved higher performance compared to the existing fault prediction system, and proposes an advanced system that can be applied to actual industrial sites.