Cognitive characteristics in firefighter wayfinding Tasks: An Eye-Tracking analysis

Abstract

During search and rescue, firefighters need to find paths in an unfamiliar space with minimum time and information available. The effective memorization and retrieval of critical spatial information can help reduce risks and increase mission efficiency. Although evidence has shown that different formats of wayfinding information, including landmarks, routes, and surveys, can impact search and rescue performance in different manners, a deeper understanding of the characteristics of firefighters’ cognitive processes related to the varying wayfinding information formats is less explored. To evaluate firefighters’ performance and cognitive characteristics in search and rescue, a firefighter experiment in Virtual Reality (VR) was conducted. Firefighters (n = 40) were recruited to participate in the simulated rescue task. After reviewing the spatial information in different formats, firefighters were requested to find three victims inside a VR maze as quickly as possible. Task performance was evaluated by the number of victims found and the time spent. Firefighters’ gaze patterns were analyzed to evaluate their cognitive status. The result showed that although the cognitive load under the survey and route conditions was significantly higher than under the landmark condition (p < 0.001), the decision-making involved a more effective cognitive process related to choosing the right path at critical waypoints such as where a turning decision must be made. Thus, the perceived workload and fatigue levels of the two conditions were lower, and the wayfinding performance was better. In contrast, with landmark information, the cognitive load levels were consistently high, along with increased mental fatigue.The findings reveal a series of cognitive features related to a more effective spatial decision-making in search and rescue. In the future, it is expected that these cognitive features can be used to develop real-time monitoring and prediction models for wayfinding performance.