London North Eastern Railway (LNER) has introduced advanced machine learning technology designed to predict train delays, following a successful trial at Peterborough and Newark Northgate stations.
The innovative tool analyses historical train performance data, passenger volumes, and weather conditions to predict potential delays, allowing station teams to proactively manage issues before they arise. The technology, accessible via LNER-issued mobile devices, informs staff of expected delays and their causes, enabling more precise support and customer service.
During the trial period, the predictive system successfully prevented over 450 potential delays, significantly improving train punctuality and reducing dwell times at trial stations.
Ian Whittles, Station Delivery Manager at Peterborough, praised the new system, noting its effectiveness in supporting rapid decision-making at busy stations. Steven Lloyd, LNER’s Machine Learning Product Lead, highlighted the additional benefits, including carbon emission reductions due to decreased idle times at platforms.
“The new capability, developed by our in-house Machine Learning Team, has proved to be an invaluable tool for my team. Peterborough is a busy station, so we are well experienced at reacting quickly and resolving issues which may occur throughout the day. The insight provided by the predictive delay tool allows us to plan more effectively, keeping our customers and our trains on the move.”
Ian Whittles, Station Delivery Manager at Peterborough
Following positive trial results, LNER is now expanding this predictive technology across its entire network to enhance reliability and customer experience.
Robin Gisby, DFTO Chief Executive, welcomed the development, emphasizing its significance in advancing railway reliability and meeting passengers’ expectations for modern, dependable travel.
