Estimating bus passenger volume based on a Wi-Fi scanner survey

Abstract

The main objective of this study is to estimate bus passenger volume based on a Wi-Fi scanner transportation survey. The study proposes a new data cleaning procedure to characterize bus passenger volume using a combination of media access control (MAC) address and global positioning system (GPS) data. We used a Wi-Fi scanner to detect the MAC addresses of individual bus passengers. The Wi-Fi scanner, used as a tool to capture passenger device data, can engage in a probe request mode to capture the MAC addresses of mobile devices or other Wi-Fi-enabled modalities without connecting to the internet. The approach proposed in this study can yield the passenger volume outputs for various bus route sections. A comparison of the passenger volume results obtained from Wi-Fi data and ground truth data indicates that the number of passengers determined from the former is less than that from the latter. The correlation between Wi-Fi estimation and ground truth is 0.78, and the trend line in both methods is similar. Therefore, the cleaning procedure proposed in this study can effectively clean raw Wi-Fi data to extract passenger volume data.