Document Type

Conference Paper


Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence


Computer Sciences, *human – machine relations

Publication Details

H-Workload 2017: The first international symposium on human mental workload, Dublin Institute of Technology, Dublin, Ireland, June 28-30.


Automated driving has been predicted to take-over from manual vehicle control in the near future. The driver’s role may then change from active operator to passive observer. Such technology offers the tantalising promise of improving driving safety. However, many studies have presented findings suggesting potentially adverse effects from automated driving systems, e.g., reduced situation awareness. Mental workload is also a key issue of concern for researchers in this area. Excessive mental workload has repeatedly been shown to be associated with degraded driving performance. Previously, most traffic psychology studies on mental workload have focused the manual driving task. However, a shift to (and from) highly automated driving will impose differing cognitive demands on the driver. For example, mental workload levels are likely to shift from underload to overload and visa-versa. Rapid resumption of manual control from a highly automated observation role seems inevitable on the basis of equipment failure or adverse conditions. Consequently, how driving performance will be effected; how it will effect driver mental workload; and how to protect road users from such system failures, are the interesting questions of concern. The aims of this experimental study are to determine the effects of control state changes (automated to manual, and manual to automated) on driver mental workload and driving performance. Participants will perform several counterbalanced driving transition scenarios (shifting between manual driving, highly automated driving and fully automated driving) in driving simulator. Dependent variables will include subjective mental workload measures, eye tracking, driving performance measures and performance on a secondary loading task. The results of this study are anticipated to provide insight into the human-machine interaction system with respect to mental workload and driving performance. Findings will contribute to our understanding of the implications of control state changes in automated driving scenarios. For example, shifting into or out of, automated driving modes. More generally, we anticipate findings could support vehicle designers by improving their understanding of the limitations of automated driving systems with respect to driver mental workload.