Causal Analysis to a Subway Accident: A Comparison of STAMP and RAIB

Accident investigation and analysis after the accident, vital to prevent the occurrence of similar accident and improve the safety of the system. Different methods led to a different understanding of the accident. In this paper, a subway accident was analysed with a systemic accident analysis model – STAMP (System-Theoretic Accident Modelling and Processes). The hierarchical safety control structure was obtained, and the system-level safety constraints were obtained, controllers of the physical layer were analysed one by one, and put forward the relevant safety requirements and constraints, the dynamic analysis of the structure of the safety control is carried out, and the targeted recommendations are pointed out. In comparison with the analysis results obtained by the Rail Accident Investigation Branch (RAIB). Some useful findings have been concluded. STAMP treats safety as a control problem and reduces or eliminates causes of the accident from the controlling perspective. Whereas RAIB obtains causes of the accident by analysing the sequence of events related to the accident and reasons of these events, then chooses one(or more)event(s) as the immediate cause and some of the key events as causal factors. RAIB analysis is based on the sequential event models, but STAMP analysis provides us with a holistic, dynamic way to control system to maintain safety.