Manchester looks to AI to ease congestion & pollution on the road

A professor and his team at the University of Huddersfield have developed an artificial intelligence (AI) system to combat congestion and curtail pollution.

The traffic management system, dubbed SimplifAI, can analyse data on the traffic flows and devise strategies for their management in a fraction of the time needed by human operators, the university said.

It will optimise timings at traffic signals for achieving the best possible flow, particularly after congestion created by unusual or unforeseen events.

Transport for Greater Manchester (TfGM) is providing testing ground for the project, which has secured more than £850,000 from the Government-backed Innovate UK.

TfGM is working in a consortium with a team from the University of Huddersfield’s School of Computing and Engineering.

Other members of the consortium on the first deployment include development consultancy KAM Futures, innovation lab FutureEverything, Human Centred Design, BT and InfoHub.

The trials are scheduled to begin later this year and due for appraisal in March of next year.

Professor Lee McCluskey, who led the university’s work on the system along with Mauro Vallati, said: “Under normal conditions, existing traffic management and traffic signals are not too bad.

“But it is very difficult for managers when normal conditions aren’t met – for example when Man United are playing at Old Trafford, or when an inner ring road link is closed due to an accident. Or there might be bottlenecks that occur relatively frequently and you want to try and alleviate them.

“Artificial intelligence is providing a tool for transport operators so they can deal with extremely complicated situations more quickly. In just a few seconds it can produce strategies composed of hundreds of different timings at traffic signals.”

The university is of view that after the trials there will be scope to market SimplifAI as a smart city system to enhance transport networks reliability and support the use of driverless vehicles.