Protection and Resilience for OLE using Computer Vision Techniques (PROLECT)

This project will utilise Computer Vision techniques applied to existing video footage and capturing a new type of video sensor to address two main challenges which are exacerbated by weather events and can result in the railway being closed. Providing means in which these type of challenge can be predicted and prevented will help provide the railway to become more resilient to weather events and season agnostic.

The following two areas will be addressed:

* Extreme hot weather causes OLE wires to extend and cause the tensioners to come into contact with the ground which can reduce tension and cause damage or event de-wirement. Utilising existing video footage this project will automatically identify OLE tensioners, position and measure them and generate an live asset map with current status level. This can be utilised as part of a digital twin model and fed into systems which are able to alert maintainers to the issue so they are able to take preventative action.
* Hot, cold and humid weather can also have an impact causing Corona discharge from electrical assets such as insulators. The Corona discharge is an early warning sign of potential damage and failure of the equipment and can be measured as part of a predictive maintenance regime to prioritise preventative maintenance activities. This project will install a UV Corona camera onto a measurement fleet or in-service train and enable automated data capture with real-time data transmission and processing. The results will be evaluated by experienced working groups to tune and amend the level of Corona events to ensure an optimum level of precision and recall ensuring an operationally useful tool.

Both of these events can have very impacts on the railway in terms of delay, safety and customer experience. By providing tools which have the capability of preventing these from occurring the railway will have an increased resilience to the weather conditions.