Pedestrians have been prioritised in a new road crossing trial using AI enabled technology in Cambridgeshire.
The Greater Cambridge Partnership has completed the trial with Starling Technologies as part of a programme to improve active travel in the East Anglian county.
The Greater Cambridge Partnership said the trial assessed whether algorithmic pedestrian crossings will reduce waiting times without impacting traffic. An algorithm understands the type of user wanting to cross through the use of three different sensors on a single unit.
“Starling funded the trial, and we supported them deploying on our infrastructure,” Dan Clarke, head of innovation and technology at Greater Cambridge Partnership, said.
“The trial is now complete, and Starling has produced a white paper with the results. We are in discussions with them on a bid into a recently announced DfT funding round for a larger deployment. Our end goal, as long as the next trial goes well, would be to deploy at scale.”
Large scale potential
A large scale deployment would allow the consolidation of the number of sensors on signal operated pedestrian crossings. This would reduce costs and increase the amount of data available to local authorities using the technology.
Existing pedestrian crossing systems prioritise vehicles and provide local authorities with little data on pedestrian usage. Each pedestrian crossing equipped with the Starling Technologies system can provide information on the type of pedestrian, the type of vehicles, near miss incidents, the density of traffic passing through, delay times and speed. The technology can identify objects and classify them and make instant decisions to benefit the pedestrian.
A deep learning model called You Only Look Once version 4 (YOLOv4) is used by the Starling Detector. Software maintenance would be carried out by Starling Technologies, but the hardware at pedestrian crossings will be the responsibility of local authorities.
Updates to the Highway Code and policy by local and national governments have placed pedestrians at the top of the travel hierarchy.
