Rinicom is pleased to announce the successful completion of the YellowLine AI project, culminating in a RIRL 5 end-to-end prototype demonstration that brings AI-driven safety support directly into the platform–train interface (PTI) environment.
The project demonstrates how real-time AI analytics can enhance situational awareness and support safer behaviour at the platform edge, one of the most safety-critical areas of railway operations.
What We Delivered
The YellowLine AI project resulted in a fully operational end-to-end testbed designed to replicate real PTI conditions.
The system integrates multiple operational data streams, including station CCTV, sensor feeds, operational logs, and public announcement channels, into a secure pipeline for real-time capture, timestamping, processing, and visualisation.
The RIRL 5 prototype uses existing station CCTV to detect PTI risk behaviours such as:
- Yellow-line crossings
- Platform-edge crowding
- Unsafe passenger positioning during train arrivals and departures
When these risks are detected, the system can trigger context-aware public safety announcements in real time, helping to reduce unsafe behaviour before incidents occur.
The prototype was successfully demonstrated during an operational showcase at Ribble Train Station (Ribble Steam Railway, Lancashire) on 27 February 2026, where industry stakeholders and end users were able to observe and evaluate the system in a realistic railway environment.
Why This Matters
The platform–train interface remains one of the highest-risk areas in railway operations. YellowLine AI aims to support operators by improving both passenger safety and operational awareness.
Passenger safety and confidence Proactive alerts help reduce near-miss situations and encourage safer passenger behaviour around the platform edge.
Operational resilience Real-time awareness of platform conditions can support staff decision-making during peak passenger loading, disruption, or crowding scenarios.
Inclusive safety by design Effective PTI risk mitigation must work for everyone, including wheelchair users and passengers who require assistance. The intervention logic within YellowLine AI has been designed with accessibility and equitable user experience in mind.
Local capability and skills development The project directly created a new role and supported an existing position within Rinicom, strengthening long-term expertise in applied AI engineering, safety assurance, and deployment readiness.
Technology Impact
YellowLine AI demonstrates a deployable real-time AI architecture designed for the operational realities of railway stations, including:
- Secure data ingestion from multiple sources
- Real-time processing and computer vision detection
- Explainable AI triggers for safety interventions
- Automated safety messaging through station systems
The project also established a controlled testbed environment, enabling iterative validation, stakeholder engagement, and reproducible evidence generation, an essential step for scaling beyond pilot deployments.
Looking Ahead
With the prototype phase now complete, Rinicom is focused on translating the results into repeatable deployment models and integration pathways for the wider rail sector.
Working with project partners, including Fujitsu, we are exploring how YellowLine AI can be integrated into broader rail service delivery and digital station infrastructure.
We welcome discussions with rail operators, safety leaders, accessibility advocates, and technology partners interested in exploring pilot deployments, safety assurance frameworks, and future rollout opportunities.
Published by E. Solera Shepherd on 12 March 2026