Esoterix to update train crowding predictions to reflect impact of delays and cancellations

Illustration of an unhappy passenger on a crowded train
Posted on: April 22, 2024 Posted by: Liz Davidson Comments: 0

Esoterix to update train crowding predictions to reflect impact of delays and cancellations

Esoterix has secured grant funding from Innovate UK, through the Bridge Ai competition, to develop Crowding Alert Technology. This extends their existing prediction technology to reflect the impact of on-the-day operational events.

Current state of the art

Crowding predictions use historic crowding to infer how busy the train will be. For example, if a service has been around 60% full over the last few weeks, it is likely to be around 60% when it next runs.

Meanwhile real-time crowding information shows how busy the train is right now. This is becoming more widely available thanks to new technologies including using Ai to process CCTV camera footage.

However, sometimes, the real-time crowding is quite different to the predicted crowding. This often happens when delays and cancellations elsewhere push people to use a different service to normal.

Crowding Alert Technology

Crowding Alert Technology will update crowding predictions across the network to reflect the current operational reality. The idea is to inform a passenger of how busy their journey will be and to do so in time for them to make alternative plans. It will also enable operators to respond sooner to unexpected surges in demand.

South Western Railway will review the work to keep it aligned with information the industry needs. Thank you to Dan Piner and the team there.

Find out more

In October 2024, Esoterix plans to deliver a demo of how the technology can help operators promptly share up to the minute crowding information to help manage demand. Get in touch to be added to the invite mailing list.

Leave a Reply:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.