EPFL Researchers Develop AI Method for Crack Detection to Enhance Railway Safety

Sep 30, 2024, 04:23 AM

Researchers at EPFL have developed an innovative method for detecting and monitoring railway track irregularities using convolutional neural networks (CNNs) and conformal prediction techniques. This advancement in artificial intelligence (AI) aims to enhance railway safety by making inspections more efficient and objective. The study utilizes high-fidelity sensor data from high-speed trains to identify cracks and other irregularities, thereby addressing a significant challenge in railway maintenance. The research was highlighted by TechXplore.