Fujitsu develop rail delay prediction solution using artificial intelligence

Fujitsu develop rail delay prediction solution using artificial intelligence

Japanese information technology giants Fujitsu are undertaking field trials to develop a new system that predicts delay time. The prediction engine is being developed in partnership with SRI International, and Jorudan Co.

Fujitsu’s new engine, which will predict the length of delays, learns from past train delay information, and delivers updates through a cloud-based service. The new function will be incorporated into Jorudan’s existing route-planner app, Norikae Annai, an app which currently offers commuters information on delays, fares, and travel times.

The prototype machine will learn from past railway operations and combine this information with train delay information provided by users of the Norikae Annai app to provide accurate, real-time predictions about delay times.

Stepping-off from the current trial, Fujitsu expect to develop a new service which embeds the time delay prediction engine. The service will utilise vast amounts of data collected from train and track sensors, and combine this with Fujitsu’s Zinrai AI technology, to enable the technology to be deployed.

(CC image courtesy of Gonmi on Flickr)

Leave a Reply

*