A Heart Rate Variability-Based Approach to Identifying Comfort and Discomfort Areas in Real-World Urban Mobility with Multimodal Sensor Data

Recently, the demand for low-speed mobility devices (LMDs), such as electric wheelchairs, has been increasing due to population aging. Despite previous efforts involving hazard mapping and subjective comfort assessment, comprehensive identification of comfort and discomfort areas using continuous physiological data has not been fully explored. Therefore, this study aims to identify comfort and discomfort areas during low-speed mobility rides and examine the environmental factors that influence passenger comfort. We propose a method to identify comfort and discomfort areas using the Heart Rate Variability (HRV) index. A data collection experiment was conducted in a real urban environment (Toyosu, Tokyo, Japan), where 37 participants rode electric wheelchairs on sidewalks. During the rides, multimodal data including HRV, GPS, environmental data, video, and survey responses were collected. Comfort levels were evaluated using the HRV index(pNN20) and visualized on a map using GPS data. The route was divided into fixed distance intervals (30, 50, and 100 meters), and each segment was assigned a comfort level based on the mean pNN20 value. Comfort and discomfort areas were defined as segments where participants consistently experienced high or low comfort. Environmental factors such as sound, temperature, humidity, and acceleration were compared between the top three high- and low- comfort areas. As a result, we found that sound levels significantly differed between high and low comfort areas at both 50-meter and 100-meter intervals, suggesting that sound may be a key factor influencing passenger comfort.