Measuring Rail Signaller Workload in a highly realistic simulated environment

The work environment of many rail signallers in Germany has been changing substantially over the past decades. In the past, rail signallers’ work used to contain many active tasks like the manual operation of switches and... [ view full abstract ]

The work environment of many rail signallers in Germany has been changing substantially over the past decades. In the past, rail signallers’ work used to contain many active tasks like the manual operation of switches and signals. Nowadays, due to increasing automation in electronic interlockings, automatic train control carries out large parts of rail signallers’ tasks during regular operations. The rail signaller has the mainly passive task of monitoring the automatic train control. During disruptions (e.g. a points failure), rail signallers face sudden intermediate periods of stressful and time critical active work. The rail signaller has to handle the disruption while at the same time ensuring safe rail operations with as little delays as possible. The long periods of passive, low-activity monitoring as well as the rapid change between passive monitoring and active, stressful work influence the workload experienced by rail signallers. The experience of a workload that is too high or too low can have significant negative impacts on task performance. Thus, it is important to analyse the impact changes in rail signallers’ work have on their workload. To assess rail signaller workload, several tools have been developed. In the UK, Network Rail’s Adapted Subjective Workload Assessment Technique (ASWAT) is used to assess the subjective workload of rail signallers using a self-report questionnaire. The present study aimed to determine whether this measure is appropriate for application in the context of rail signalling in German railway operations. The study was conducted by human factors experts of the DLR at the Railway Operation and Experimental Laboratory of the Technische Universität Berlin using their simulation of a rail signallers work environment in an electronic interlocking. Eleven rail signallers took part in the study. Participants completed three different experimental scenarios in a randomized order, one scenario with automatic train control, one without and one scenario in which a sudden points failure occurred. Thus, the experimental design allowed a comparison between the workload during passive work, active work and during the handling of a disruption. We assessed subjective workload using a German translation of the ASWAT as well as the NASA Task Load Index. In addition, we used behaviour coding to gain an understanding of the objective workload during the scenarios. This study is a first approach to validate the use of the ASWAT in the context of rail signalling in German railway operations. A successful use of the ASWAT in this context will allow the comparison of results regarding rail signaller workload cross-culturally. Thus, it would be possible to compare e.g. the impact of automation on rail signaller workload across different operational systems and concepts of automation. The results could be used to improve rail signallers’ work environments in terms of their user friendliness, adaptive automation and interface design.

Note: We would like our paper to be considered for inclusion in a special edition of Applied Ergonomics in addition to being published in the electronic book of conference proceedings.