Detection of high-speed red-aspect approaches using a multi-data approach

This paper proposes a proactive safety indicator for Signals Passed at Danger (SPAD) using multiple data sources. The proposed technique integrates data sources using a graph database and R software to store, process and analyze the train services and signals data to yield a key performance indicator for high-speed red-aspect approaches. The method is illustrated using a case study where three data sources were used viz., On Train Data Recorders (OTDR), Red Aspect Approach to Signals (RAATS) data and the railway infrastructure manager's signal database. The proposed approach aims to shift safety management for SPADs from 'avoiding things going wrong' to 'ensuring that everything goes right'. The approach is complementary to the current driver competency performance system, potentially allowing the driver supervisors to evaluate drivers' braking style without any subjectivity. It also indicates which signals on the infrastructure may be particularly prone to SPADs, even if a SPAD has not occurred there yet.