Driver performance modelling and its practical application to railway safety.

Abstract

This paper reports on the development and main features of a model of driver information processing. The work was conducted on behalf of Network Rail to meet a requirement to understand and manage the driver's interaction with the infrastructure through lineside reminder appliances. The model utilises cognitive theory and modelling techniques to describe driver performance in relation to infrastructure features and operational conditions. The model is capable of predicting the performance time, workload and error consequences of different operational conditions. The utility of the model is demonstrated through reports of its application to the following studies: Research on the effect of line speed on driver interaction with signals and signs. Calculation of minimum reading times for signals. Development of a human factors signals passed at danger (SPAD) hazard checklist, and a method to resolve conflicts between signal sighting solutions. Research on the demands imposed on drivers by European train control system (ETCS) driving in a UK context. The paper also reports on a validation of the model's utility as a tool for assessing cab and infrastructure drivability.