Cognitive interference during driving

In his dissertation, Moritz Held investigates how multiple cognitive concepts such as perception, attention, memory, and motor control interact under mental load, using driving as a realistic and safety-critical example. Because driving consists of many subtasks, simple observation of behavior alone is often insufficient to fully explain what happens in the driver’s brain. In his research, Held therefore addresses this issue by using a multi-method approach employing computer simulation, behavioural observation, eye-tracking and neuroimaging, in our case measuring the activation of brain areas.

Held uses eye-tracking and pupil dilation as indicators of attention and cognitive load to investigate where drivers are looking and how mentally exhausted they are at that moment. In addition, he simulates the mental processes using computer software to understand the individual components of complex actions, as well as predict future driving actions. Functional Magnetic Resonance Imaging (fMRI) is a method to investigate the brain activation and is used in Held's thesis to measure the activation of certain brain regions in complex driving situations.

The results show that both high and low mental load can impair driving performance. High load leads to competition for limited brain resources, whereas very low load promotes mind-wandering and a lack of focus towards driving, which also degrades performance. Cognitive modeling shows how targeted interventions can mitigate these negative effects, provided the interventions pose minimal additional cognitive demands. In complex scenarios such as left turns at intersections, a clear sequence of brain activation was observed, consistent with theoretical psychological models.

The dissertation suggests that mental load can be monitored and regulated using a combination of computer simulation and brain signals. These findings provide a foundation for adaptive driver-assistance systems that adjust support based on the driver’s mental state and are also applicable to other safety-critical domains, including aviation, power-grid control, and human–robot interaction.

Read more: Driving while distracted: 'Phew, I finally made it’