Pedestrian traffic volume on the frontage road is regarded as a crucial factor influencing customer inflow in retail location strategies. Therefore, this study aims to develop a method for accurately estimating pedestrian traffic volume. While conventional visual surveys have difficulty covering small-scale streets, the use of GPS location data obtained from smartphones enables the measurement of pedestrian flows across a wide area, including alleys and residential streets. A key novelty of this study lies in incorporating the accuracy information (mean error radius) attached to each GPS point into the map-matching process, thereby reflecting point-specific error characteristics. As a result, the proposed method allows for the estimation of traffic volumes that account for sidewalks on each side of arterial roads and provides multiple aggregation patterns that correct for biases in the number of observation days per user. These findings contribute to advancing analyses of commercial agglomeration structures and improving land-use models for retail location planning.
Watanabe et al. (Thu,) studied this question.