◆ 참여자: 유태경, 조재윤, Gazikhodjaev Nodirkhuja, Jamwal Harishik Dev Singh, Chaulagai Nishant

◆ 개발기간: 2024.9~2024.12

◆ 프로젝트 내용: This project was started to reduce derailment accidents, which are one of the most serious train accidents. Derailments during train operations are not only technically and financially damaging but can also lead to severe loss of lives. The ultimate goal of this project is to predict defects in point machines, which are closely related to derailments.

The project aims to develop an AI-based model that detects and predicts errors in point machines, providing maintenance workers with early warnings about potential risks. By learning from large amounts of current and voltage data from point machines, the model is designed to distinguish between normal and abnormal states, analyze trends in abnormal data, and predict future errors.

The goal of this AI system is to detect whether the data from the point machine is normal or abnormal and predict when abnormal data might occur. This can help reduce potential train accidents in the future and be applied in various ways to improve the railway industry.