Cognition-driven navigation assistive system for emergency indoor wayfinding (CogDNA): Proof of concept and evidence

Abstract

Attributed to the advances in sensing and visualization technologies, novel wayfinding assistive systems are becoming more accessible such as the augmented reality (AR)-based wayfinding guidance that provides immersive information about the space and environments. However, the excessive information provided by these new methods results in cognitive overload, leading to subpar performance. Especially for emergency wayfinding where responders need to process a large amount of information while performing search and hazard identification tasks, an overdesigned wayfinding user interface and excessive information can cause confusion or disorientation among responders. This study proposes and tests a real-time cognition-driven navigation assistive system for emergency wayfinding called CogDNA. CogDNA measures responders’ real-time cognitive load and mental status based on the high-frequent pupillometer and gazes tracking data. Then the elements and formats of the primary wayfinding information are adjusted based on the individualized cognitive load models and the mental fatigue status measured by the pupillometer data including gaze movement entropy and blink rate. To test the efficacy of CogDNA, a human-subject experiment (n = 31) was performed with local firefighters in a simulated emergency wayfinding task. Three conditions were presented including the control condition with a static wayfinding system, a self-adaption condition where responders could adjust information with gesture controls, and the auto-adaption condition that tailored information based on real-time cognitive load and mental status measures. The result shows that the proposed method reduces the workload and improves the performance of the responders during the wayfinding task. The self-reported assessments also indicate benefits of the proposed method in cognitive load. The findings prove the efficacy of adaptive wayfinding information systems based on real-time cognitive load measures for future emergency wayfinding tasks.