In Canada, thermal stress buildup due to climatic changes can lead to failure of rail structures. Currently, there are no techniques available to continuously monitor the stress state in a rail over long distances. Distributed fibre optic sensing is a promising solution. Due to the extent of rail track infrastructure in Canada (45,000 km), the method of sensor install must be automated. In addition, to manage the immense volumes of measurement data as well as implement predicative analytical strategies, artificial intelligence (AI) approaches are the essential.
The proposed research addresses these needs through a combination of lab and field research, including the development of AI-based methods, to account for temperature and detect critical rail stresses, and a robotic fibre installation prototype. The project outcome will be a prototype digital twin that can eventually be used by industry collaborators for use in the maintenance and management of their rail networks.