Accidents are caused by plural human errors, each of which is influenced by multiple factors. Therefore, there are a number of error-prevention measures that can be considered from a single accident. To effectively prevent accidents on a limited budget for a limited period, preventive measures will therefore be prioritized. To support judgment in this process, the RTRI has established a technique to evaluate human-error risk in train operation.

The method first calculates a risk value from error frequency and the maximum scale of an accident that might result from a particular error. To prevent errors adequately, it is necessary not only to identify high-risk events but also to consider risk-inducing factors. This technique therefore combines effects of the risk value and risk-inducing factors to quantitatively determine the priority of error prevention measures (Fig. 1).

Figure 2 shows a case in which the technique is applied to a train operation depot. The analysis shows that appropriate education and training are required on the scene where the conditions change ((10) in Fig. 2) as errors tend to occur when the working environment or the details of work change. The analysis also shows that the introduction of an alarm or abnormality detection system is effective to support drivers, who sometimes fail to notice abnormalities ((3) in Fig. 2).

The RTRI has developed a manual summarizing the requirements and procedures for risk evaluation. This can be used to identify effective error-prevention measures at different depots by inputting error events and their specific causes.