Improved Energy Efficiency of Railway Transport using AI

(AI Factory /RAILWAY)

Goal

The main goal of this project is to develop and demonstrate an integrated approach to condition monitoring of track using AI and digital technologies.

Project status and results

Trust in Swedish railway transport has recently declined, leading to a shift towards the less energy-efficient road transport. Track condition significantly impacts capacity and punctuality, which are crucial factors for railway reliability.

This project aims to improve track conditions and encourage the return of traffic to the more energy-efficient rail option. A key challenge in track defect detection using vibration measurements is differentiating between regular design elements (like turnouts and joints) and actual defects.

This project focuses on leveraging AI techniques to identify these planned elements and detect defects in their immediate vicinity. This involves data labeling, signal processing, and anomaly detection. Additionally, the project will explore solutions that combine sensor fusion with other data sources, such as LIDAR measurements and satellite data. Furthermore, the project will develop a demonstrator as a proof-of-concept using the existing "AI Factory for Railway" (AIFR) framework developed by Luleå University of Technology.