Spiral tunnel driving and cognitive load: An eye-tracking investigation into tunnel geometry and traversal effects.

Abstract

Objectives: Despite the increasing construction of spiral tunnels, the specific scientific problem of how their unique geometric characteristics (length and radius) and travel direction (upward vs. downward) collectively influence drivers' cognitive load remains insufficiently understood, with a lack of systematic quantification using eye movement metrics. This study aimed to evaluate and quantify the cognitive load experienced by drivers navigating spiral tunnels, focusing on addressing this gap by examining the effects of tunnel geometry and travel direction through eye movement metrics.

Methods: A naturalistic driving experiment was conducted with 30 licensed drivers in three spiral tunnels varying in length (1,330, 2,200, and 4,460 m) and radius (1,000, 850, and 700 m). Eye movement data, including fixation duration, pupil diameter, saccade duration, and saccade amplitude, were collected and analyzed to assess cognitive load.

Results: Increased tunnel length and decreased radius have been associated with greater cognitive load. Specifically, the average fixation duration in the 4,460-m-long, 700-m-radius Hankou Tunnel is 4.6% higher than that in the 2,200-m-long, 850-m-radius Liuyuan Tunnel and 12.9% higher than in the 1,330-m-long, 1,000-m-radius Nanping Tunnel. The average pupil diameter in the Hankou Tunnel is 3.5% larger than that in the Liuyuan Tunnel and 7.7% larger than in the Nanping Tunnel. The average saccade duration in the Hankou Tunnel is 14.8% longer than that in the Liuyuan Tunnel and 34.0% longer than in the Nanping Tunnel, while the average saccade amplitude in the Hankou Tunnel is 5.0% smaller than that in the Liuyuan Tunnel and 14.0% smaller than in the Nanping Tunnel. Drivers have also experienced higher cognitive load during upward traversal compared to downward traversal, with the average fixation duration during upward traversal being 10.0% higher than that during downward traversal, the average pupil diameter during upward traversal being 4.0% larger than that during downward traversal, the average saccade duration during upward traversal being 10.6% longer than that during downward traversal, and the average saccade amplitude during upward traversal being 5.5% smaller than that during downward traversal. No significant interaction effects have been observed among tunnel length, radius, and travel direction on the eye movement metrics.

Conclusions: Optimizing tunnel design is crucial to minimizing cognitive demands on drivers. Shorter and wider tunnels are recommended, and design features such as enhanced lighting and improved signage should be considered to mitigate the additional cognitive load of uphill driving. These findings have significant implications for enhancing driver safety and performance in spiral tunnels.