Autonomous maintenance of infrastructure, rolling stock and interfaces

Algorithms to detect and quantify defects in an S&C using on-site measurement data.

Goal

Develop algorithms to detect and quantify defects in an S&C using on-site measurement data.

Project status and results

The project builds on previous research results and data obtained from DigiSwitch, at Luleå University of Technology. The project aims to design algorithms, approaches, techniques and methods based on using feature extraction and artificial intelligence to perform anomaly detection on railway S&C. The methods and concepts that the project will be based on are: machine learning, signal processing and vibration sensors.

Sponsor: Trafikverket
Researchers: Matti Rantatalo (PL), Jan Lundberg, Taoufik Najeh, Yang Zou (PhD student)
Duration: 2021-22