THE MANAGEMENT OF UNCERTAINTY: MODEL FOR EVALUATION OF HUMAN ERROR PROBABILITY IN RAILWAY SYSTEM

Abstract

The methodology of quantitative risk assessment in use in the railway industry refers the RAMS approach, as expected in EN 50126, to define the failure of the system. From the probability of failure follows the evaluation of the failure consequence, the check of the accident occurrence and finally, the analysis of the achievable accident scenarios to quantify the damage as result thereof. In this approach, the central concept is the failure phenomenon, i.e., the deviation of the proper performance of a function by a system component. The case-study is referred to the maintenance procedure, necessary to ensure safe use of the railway infrastructure. The homogeneous treatment of human behaviour in the proposed risk model involves the definition of the function of reliability (or failure) of "human component" since the maintenance procedure of railway system shows a high probability of human error. The paper analyzes a series of control checks for back analysis testing of accident occurrences and highlights the opportunity to perform quantitative risk analysis comparing alternative designs of railway system by analogical analysis methods. In this analysis the evaluation of human error according to combined techniques of risk analysis and Ishikawa’s theory is relevant because in the railway industry organized procedures of good practice usually have been employed.