Inventor: Vladimir Vilde, Department of Engineering and Centre for Digital Built Britain
Mentor: Martin Carroll
It is estimated that there are 3000 sub-standard bridges in the United Kingdom at any one time. Bridge repairs can be very disruptive and costly, requiring closures which can have significant impacts on road and rail traffic.
Currently there are no good and accessible tools to monitor bridge status on an ongoing basis. Although there are a number of sensor-based systems, these can very expensive to install on a single bridge, on top of which monitoring and service costs must be added. This often works out as being more expensive than just waiting until a repair is needed, especially if the repair is seen as non-urgent.
The inventor is working on ways to address this issue and improve the long-term monitoring of bridges and other infrastructure in a cost-effective way. Building on his PhD research on artwork monitoring for galleries and museums, he has developed an algorithm which can analyse images of bridges and highlight any relevant changes.
The algorithm, which is initially being developed for bridges, uses the photos which are typically taken as part of a formal survey. The aim is to gather additional sources of data and further develop it into an automatic tool which can assist the surveyors in their role, rather than trying to replace the need for expert surveys.
The challenge for the i-Team is to investigate other possible uses of this approach. What other types of infrastructure could benefit from an automatic analysis of images, taken either by expert surveyors or members of the public, and identify the changes which indicate that repairs are needed? Examples could include sports fields and equipment, rental cars, AirBnB and long-term rental properties, or even offshore structures and wind turbines.