Task-induced fatigue when implementing high grades of railway automation

Abstract

The study was focused on the effects of different grades of railway automation on task-induced fatigue and workload in train drivers and, when considering high grades of automation, operational staff in a control centre, so-called train operators. Train operators remotely monitor and manually drive automated trains upon system request during disruptions. As the task environment substantially differs depending on the grade of automation, effects on task- induced fatigue and workload levels were expected. To quantify and compare these effects, a simulator study with professional train drivers (N = 32) was conducted and the grade of automation was manipulated experimentally between subjects according to the railway specific automation taxonomy. Fatigue was assessed by the Karolinska Sleepiness Scale prior to and after a simulated working period of 2 h. Workload was assessed using the NASA-TLX. The results showed (a) significantly increasing fatigue levels over time (b) significantly higher fatigue ratings as a result of working with an intermediate grade of automation in comparison to working with a high grade of automation. Workload scores were (c) significantly higher in the high grade of automation group. Consequently, the introduction of high grades of automation may be used to tackle longstanding fatigue issues associated with train driving. The role of workload and limitations of the current research are discussed and recommendations for future research and operating companies are provided.