Eye Tracking While Multitasking and Driving

This case study examines the research presented in “Multitasking while driving: Central bottleneck or problem state interference?” by Held, Rieger, and Borst, highlighting the critical role of eye tracking in understanding cognitive workload and driver behavior. The study investigates how visuospatial attention and working memory load interact during driving, specifically addressing whether interference occurs at a central control resource or a task-specific information processing resource.

Previous research on multitasking while driving has observed interactions between cognitive concepts like attention and working memory, often attributed to a “central bottleneck” or a “problem-state bottleneck” related to working memory usage. To differentiate between these hypotheses, Held, Rieger, and Borst developed two cognitive models within the ACT-R architecture, implementing each bottleneck theory. They then conducted an experiment with human participants to validate these models, varying visuospatial attention and working memory load during a simulated driving task.

Multitasking and Eye-Tracking Research Methodology

A key component of the human experiment was the use of a EyeLink Portable Duo eye-tracker. This technology allowed for the precise measurement of gaze position and pupil dilation at a sampling rate of 500 Hz, without the need for head-stabilization. The researchers utilized eye-tracking data in two primary ways:

1. Assessing Cognitive Workload: Pupil dilation is a well-established indicator of cognitive workload. The study analyzed changes in pupil size relative to a baseline, confirming that increased n-back difficulty (a measure of working memory load) led to a significant increase in pupil size. This validated the experimental manipulation of working memory load, demonstrating that participants were indeed experiencing higher cognitive demands.
2. Monitoring Attention Allocation: The eye-tracking data also revealed a significant decrease in fixations on the speedometer with increasing n-back levels. This finding provided behavioral evidence that as the working memory task became more demanding, drivers allocated less visual attention to critical driving instruments, indicating a diversion of resources from the primary driving task.

Eye-Tracking Results Show Increased Cognitive Load Reduces Visual Monitoring

The study found that the problem-state-bottleneck model more accurately accounted for decreased driving performance due to working memory load and increased visuospatial attentional demands compared to the central-bottleneck model. The eye-tracking results directly supported this conclusion, showing that increased cognitive load (as indicated by pupil dilation) led to a reduction in crucial visual monitoring of the speedometer. This suggested that as working memory load increased, drivers performed fewer control actions, resulting in decreased driving performance.

The eye tracking data offer valuable insights for the development of adaptive automation systems in vehicles. By quantifying the contribution of different mental loads, such systems could potentially monitor a driver’s cognitive state (e.g., through physiological sensors like eye-tracking) and adapt driving responsibilities to alleviate cognitive workload when necessary. This could lead to substantial improvements in traffic safety by ensuring that drivers remain attentive and engaged, with assistance provided when their capabilities are stretched. The ability to observe and quantify real-time attentional shifts and cognitive effort through eye tracking was instrumental in validating the models and advancing this understanding.

For information regarding how eye tracking can help your research, check out our solutions and product pages or contact us. We are happy to help!