10168917AI-enabled rail clash alert early warning systemClosedProcurementInnovate UK**Overview** JNCTION is trialing an AI-enabled early warning system designed to improve the operational efficiency, sustainability, and predictability of the UK rail freight sector. The system predicts and prevents potential clashes between freight and passenger services, enabling freight operators to make proactive, data-driven decisions. This innovation directly supports Innovate UK objective of "mitigating harmful emissions at source" by avoiding diesel trains having to repeatedly stop and start between planned stops. **Key Benefits** * **Emissions Reduction**: Each 'line speed to stationary' operational clash increases journey emissions by approximately 5%, based on ongoing RSSB research. By reducing the frequency of such clashes, the clash alert system supports the UK's decarbonisation targets. We estimate a £537,590 per year benefit from reduced NOx, PM and CO2 using the TAG Air Quality valuation methodology. * **Operational Efficiency**: In 2024/25, the majority of freight operator delay minutes and a high number of passenger delay minutes were due to operational issues. The system provides short (15 min), medium (30 min), and long (1 hour) horizon alerts, allowing control rooms to prevent delays before they occur. * **Digital Integration**: Using real-time tracking and historic data, the system presents alerts through a tested dashboard interface. This enhances coordination with other logistics nodes, such as ports and distribution centres, enabling smarter multimodal operations. * **Modal Shift Enablement**: High unpredictability in passenger and freight services is a key barrier to modal shift from road to rail. By increasing reliability, the system supports this modal shift---crucial for reducing CO2, congestion, and heavy vehicle reliance. One passenger train can remove up to 500 cars from the road, whilst one freight train can replace up to 110 lorries, producing 76% less CO2 per tonne than road freight. * **Network Decongestion**: Reduced clashes free up capacity across the rail network, benefitting both passenger and freight services. **AI-enabled** * JNCTION's solution leverages AI to identify clashing paths on our digital twin of the network * We build upon our existing AI ensemble model approach to blend deterministic logic, statistical inference and ML prediction at different time horizons, depending on the most reliable weightings for the context.