Development of a prototyp for trainbased detection of track defects
|Coordinator||Luleå tekniska universitet - Avdelningen för Drift, underhåll och akustik|
|Funding from Vinnova||SEK 2 180 000|
|Project duration||April 2018 - December 2019|
|Venture||The strategic innovation programme InfraSweden2030|
Purpose and goal
This project aimed to create a system where measurements and condition monitoring of railway fasteners and rails can be carried out using in service trains. The goal of the project has been to construct a prototype for train-based condition monitoring of rails and fasteners. This goal has been broken down to the following sub-goals: 1 Development of two prototypes 2 Verification of prototype in the field and development / adaptation of detection algorithm 3 Verification of the sensor by means of a video-equipped two-way vehicle 4 Plan for mounting the sensor on trains in operation
Expected results and effects
The project has resulted in a solution and prototypes for condition monitoring of railway fasteners ready and prepared for field tests on trains. The project result has met the initially set expectations where a completely new electronics design has been developed where circuit boards have been manufactured with associated software for eddy current measurement. Validation measurements using a line camera system have been performed to compare with image recordings of the same rail. Manual measurements of healthy and defected fastening systems have also been performed.
Planned approach and implementation
The project has been carried out mainly by performing three parallel activities. 1 - Development of electronic solution and hardware such as antennas etc. 2- Adaptation of detection algorithm and machine learning methods. 3 - field test for sensor and detection algorithm validation. In the project, Bombardier has been responsible for the electronics development and Luleå University of Technology has been responsible for algorithm development and field validation as well as the creation of validation data with linked visual measurements together with Raildoc.