Japanese company Odakyu Electric Railway is trailing Nokia’s artificial intelligence (AI)-based SpaceTime scene analytics for railway crossing safety.

Running until next month, trials have already begun at Tamagawa Gakuenmae No.8 railroad crossing in Machida City, Tokyo. These tests will identify potential improvements for rail crossing safety.

Nokia Bell Labs developed SpaceTime scene analytics. It applies machine-learning-based AI to available camera images and can detect abnormal events.

It also analyses available image feeds generated by conventional railway crossing cameras. This allows it to identify potential issues in real-time.

Running on edge computing resources, it minimises the required bandwidth at remote sites with limited connectivity.

Nokia Japan head John Harrington said: “Odakyu Electric Railway is renowned for being an early adopter of new technology and this trial illustrates the role that AI can play in delivering enhanced levels of vigilance.”

SpaceTime scene analytics can provide real-time alerts for unauthorised entry into remote facilities.

It can also detect and alert supervisors when personnel or equipment access unsafe locations, or when heavy machinery comes out of position.

Harrington added: “Network-connected cameras are one of the most prolific sources of internet of things (IoT) data that can provide valuable insights to help promote high safety standards.

“By running machine learning analytics on camera feeds and sending solely relevant scenes and events to operators, the full benefits of video surveillance can be realised in a wide variety of settings, with rail crossings a particularly relevant use case.”

Odakyu Electric Railway currently has 229 crossing points across 120.5km of rail track and 137 radar systems to detect objects.

In December, German railway company Deutsche Bahn (DB) selected Nokia to test standalone (SA) 5G technology for automated rail operations.