Northern uses artificial intelligence to count the number of passengers on its trains
Train operator Northern is using artificial intelligence (AI) to count the number of passengers on some of its services.
The train operator has deployed an Automatic Passenger Counting (APC) algorithm – which tests show is over 98 per cent accurate – to software on its fleet of eight Class 769 trains.
The work, which was part-funded by Rail North Partnership, was carried out at Allerton TrainCare Centre in Merseyside.
It saw a team of 15 specialist engineers from Northern and its technology partners Icomera and DG8 install 88 cameras and 8 computers using 5km of cables.
The Class 769 fleet, which is based at Allerton, operates services between Liverpool, Manchester, Southport, Wigan and Blackpool.
Jack Commandeur, engineering director at Northern, said: “More than 90m people step on-board one of our trains every year and our fleet of Class 769 trains will now give us more accurate data on passenger volumes than we’ve ever had before.
“This joint collaboration with Icomera and DG8 will help us to improve the overall efficiency of our operations and can inform future service planning.”
Peter Kingsland, Icomera’s senior vice president in the UK, said: “The on-board network on the newly upgraded Class 769 fleet provides Northern with an exciting testbed for future applications of AI video analytics.
“As the rail industry’s digital transition continues to accelerate, these applications can be combined with the passenger services we support to make every journey safer and more time-efficient for passengers, as well as more cost-effective for rail operators.”
As part of Northern’s new ‘Plan Led Approach’ to engineering, the work was carried out alongside scheduled heater upgrades and improvements to WiFi and CCTV equipment to ensure each train was out of passenger service for the minimal amount of time.
Northern is the second largest train operator in the UK, with 2,500 services a day to more than 500 stations across the North of England.
