Detecting cognitive underload in train driving: A physiological approach

Some train driving situations can be very monotonous (eg running at low speeds for a long time) or repetitive (eg running on caution signals or driving routes with frequent station stops). There is also increasing automation... [ view full abstract ]

Some train driving situations can be very monotonous (eg running at low speeds for a long time) or repetitive (eg running on caution signals or driving routes with frequent station stops). There is also increasing automation in the rail industry.
With these issues in mind, there is an emerging concern that the train driving task may not demand sufficient attention from the driver to keep them alert and engaged. Situations involving very low task demands may lead to cognitive underload, which occurs when the demands of a task are so low that the performance of the person carrying out the task suffers. Cognitive underload is thought to be associated with feelings of boredom, monotony, low motivation, fatigue, sleepiness, loss of attention and impaired ability to deal with unexpected situations. This may result in drivers missing critical information required to maintain the safety of the train or the optimal operation of the network.
This exploratory study aimed to understand whether it is possible to detect and measure the effects of cognitive underload in train drivers. Fifteen train driver participants completed an extended version of the Sustained Attention to Response Task (SART). Participants’ performance on the SART was recorded and cardiac and respiration activity as well as skin conductance were monitored. It was found that participants sometimes entered an automatic response mode. Responses tended to be quicker and their heart rate was lower just before they made an error. This is a new finding and means that this is a promising area to study. It was predicted that task performance would deteriorate over time but this was not observed.
The outcome of this study has implications for the implementation of measures which may help to mitigate the deleterious effects of cognitive underload in train drivers, as well as for the design of future studies on cognitive underload. The results give an important insight into the relationship between task performance and physiology during low workload conditions. If physiology can be shown to reliably predict errors due to underload, physiological indicators could be used to identify when drivers are experiencing cognitive underload, and an appropriate mitigation could be applied. Future work in this area, therefore, has the potential to provide real benefits to the operation of the railway, and we would encourage the rail industry’s participation in future studies on cognitive underload. Studies in this area should also use longer trial periods and naturalistic study designs, which may become possible with the development of accurate and non-invasive technology for physiological monitoring.
We would like this paper to be considered for inclusion in a special edition of the Journal of Rapid Rail and Transit.