Visual Occlusion and Attentional Demand in Driving¶

Status: established
Last updated: 2026-05-31
Sources: Kircher 2021 Visual Occlusion Attentional Demand
Tags: [eye-tracking, driving, occlusion, attentional-demand, spare-capacity, visual-demand, distraction]

Summary¶

Visual occlusion is a method in which a driver's forward vision is temporarily blocked to measure how long they can safely suspend visual sampling of the road environment. Kircher et al. (2021) synthesised findings from four occlusion studies demonstrating that drivers possess variable visual spare capacity, modulated by situational complexity. The method provides a direct measure of attentional demand that does not require a secondary task and is distinct from glance-based eye tracking metrics.

Body¶

Context¶

Kircher et al. (2021) examine visual occlusion as a tool for assessing attentional demand and spare capacity in driving, compiling and discussing findings from a series of four of their own occlusion studies — a field study with an instrumented vehicle, a fixed-base simulator, a motion-base simulator, and a linear motion-base platform — in which a driver's forward vision is temporarily blocked to measure how long visual sampling of the road can be suspended without incident. The text is a four-page conference abstract (DDI 2021 Book of Abstracts), so locators below are given to the abstract's own pages. Within this knowledge base the article is the driving-attention applied case, sharing the attentional-demand and spare-capacity theme with the workload strand in Pupil Dilation Cognitive Load and Eye Tracking In Surgery and the involuntary-sampling theme with Gaze Based Hci And Usability; it offers an alternative to gaze-based demand metrics by measuring tolerated time without vision rather than where the eyes land.

Key Points¶

Visual occlusion blocks the forward roadway to measure how long drivers tolerate interrupted vision; the four studies used a range of apparatus, including a micro-switch-operated device behind the steering wheel and screen occlusion, with peripheral vision either preserved or removed. In self-paced paradigms drivers control occlusion onset and duration; in fixed-duration paradigms the duration is set, and one study used an irrevocable system-paced occlusion that continued until a crash or run-off-road (PDF p. 3, orig. p. 36). The approach was pioneered by Senders et al. (1967), who used it in real traffic to assess the visual demands of different traffic environments (PDF p. 2, orig. p. 35).

Across the four studies the central finding is that drivers possess visual spare capacity — the possibility of executing an additional visual task while driving — to varying degrees, shown by their ability to occlude without ensuing incidents or collisions. Situations of higher prediction uncertainty, such as oncoming traffic or intersections, lead to less frequent and shorter occlusions, while monotonous environments such as motorways produce longer occlusions and prior occlusion history predicts future occlusions, indicating adaptive sampling (PDF p. 3, orig. p. 36). Many participants in Study I reported occluding below their maximum capacity, keeping a safety margin, and this is supported by Study IV, where collisions or run-off-roads occurred on average only after twice or more the typical self-paced occlusion duration (PDF p. 4, orig. p. 37).

The theoretical contribution is to frame gaze during driving as predictive sampling rather than continuous scanning: drivers sample so as to predict and prepare for the near future. In Study II the likelihood of occluding was lowest on approaching an intersection and while closing the gap to an oncoming vehicle, and increased once past the first half of the intersection or once the oncoming vehicle was near; in Study IV, under irrevocable occlusion, drivers corrected their trajectory using information sampled before the occlusion began (PDF p. 3, orig. p. 36).

The primary limitation is the absence of a benchmark: no concrete criteria fix whether a given occlusion frequency and duration is below, at, or over capacity, since an incident clearly signals insufficient sampling but its absence does not guarantee that enough was sampled (PDF p. 4, orig. p. 37). Operating the occlusion apparatus by hand or foot may itself require more mental effort than the more natural closing of the eyes, although the authors note that closing the eyes leads quickly to fatigue and lacks the flexibility of an external device (PDF p. 4, orig. p. 37).

Conclusion¶

Kircher et al. (2021) conclude that visual occlusion gives a direct measure of attentional demand without a secondary task, and that drivers' spare capacity is real but situation-dependent and managed through predictive, anticipatory sampling rather than continuous scanning. The method's value as a demand metric is tempered by its lack of an absolute benchmark for adequate sampling and by the attentional cost of the apparatus itself.

References¶

Kircher, K., Ahlström, C., Kujala, T. & Liu, Z. (2021) 'Visual occlusion as tool to assess attentional demand and spare capacity', in DDI 2021: 7th International Conference on Driver Distraction and Inattention, Book of Abstracts, pp. 35–38. Université Gustave Eiffel; SAFER; UNSW Sydney. kircher2021occlusion

Senders, J. W., Kristofferson, A. B., Levison, W. H., Dietrich, C. W. & Ward, J. L. (1967) 'The attentional demand of automobile driving', Highway Research Record, 195, pp. 15–33. To be validated. kircher2021occlusion

Open Questions¶

Can the predictive sampling model from driving be applied to operators in remote monitoring or control room contexts?
How does occlusion tolerance change with automation level — do highly automated vehicles reduce situational demand sufficiently to increase spare capacity?
Is the safety margin in self-paced occlusion consistent across age groups and experience levels?