Project Summary

Headway Rail Innovations Ltd AI-enabled rail freight clash alert early warning system will combine historic train movement observations with real-time feeds to accurately predict short (under 15 minutes), medium (15-30 minute) and long (30-60 minute) time horizon clashes of rail freight with other rail services.? Information and alerts will be displayed on a user-friendly dashboard, enabling control teams to reduce clashes, reduce congestion, and reduce emissions, as well as improve efficiency and encourage freight shift from road to rail. Headway Rail Innovations Ltd has extensive experience working on delay predictions, network optimisation and emissions modelling.

Project Achievements

The project successfully progressed from an experimental proof of concept (TRL 3) to a small- scale prototype validated in a testing environment (TRL 4). This was determined by the delivery of a working MVP that integrates live operational data (TRUST/CIF) with an ensemble machine learning model. Key results include demonstrating that our probabilistic approach provides improved predictive accuracy over traditional deterministic baselines when tested against historic movements.

Conclusions

The project contributes to the field by shifting rail conflict management from reactive, deterministic modelling to proactive, probabilistic forecasting. A significant finding was that historic lateness is the strongest individual predictor of future train performance, outperforming more complex pathing calculations. The project was highly successful, meeting all core objectives and receiving formal validation from stakeholders at Heavy Haul Rail, Chiltern Railways, and the Elizabeth line

Next Steps

The project has successfully advanced to TRL 4, and our immediate priority is to have a live operational pilot (TRL 5/6). This next phase will focus on measuring real benefits, specifically emissions reductions and delay mitigation, under real-world conditions. We aim to implement this live trial within the next 6–12 months, ideally expanding beyond the Chiltern geography to other high-complexity mixed-traffic routes.