Towards remote supervision and recovery of automated railway systems: The staff’s changing contribution to system resilience

Rail automation in mainline operation raises questions about system resilience, becauseit goes along with increased technological and operational complexity, while ultimately discarding the train driver as front-line staff situated onboard the train. We argue in favor of keeping the layer of resilience associated with the train driver in the system through a train operator workplace that enables remote supervision, diagnosis and intervention of automated rolling stock out of a control center. Firstly, results from a simulator study featuring a prototypical remote workplace, developed in an iterative user-centered design process, show positive feedback in a sample of active German train drivers (n}=textbf{{20) in terms of acceptance ratings, usability assessment and perceived benefit to system resilience. Secondly, lower than optimal workload ratings in two experimental conditions of varying task load indicated free capacity for additional operational tasks throughout periods without direct manual interventions. Further insights on task load and workplace design as well as their limits are discussed, pointing out the direction for the next steps to be taken in the following design iterations. Future research needs to generate measures suitable for comparing system resilience between different grades of automation.