A Method for Estimating the Time of Pedestrians Encounter Timing by Measuring Bluetooth Signal Strength with Multiple Devices

Traffic safety support systems detect and warn of potential accidents by tracking the location of pedestrians and vehicles. However, most pedestrians and vehicles that do not voluntarily provide location information to the systems will not be tracked by the systems. Therefore, to make the system more advanced, it is necessary to acquire the positions of these pedestrians and vehicles and to predict their positions in a few seconds. This challenge can be partially addressed by having pedestrians who participate in the systems detect the passing of pedestrians who are not tracked by the system and report the passing location and time to the systems. The approximate location of untracked pedestrian is the position of the system participant at the time of passage. Therefore, it is important to obtain the time of passage. The purpose of this research is to estimate the time of passing from the Bluetooth RSSI time series data. This paper proposes a machine learning-based method for estimating the time that a pedestrian wearing a Bluetooth receiver at five points on the body passes a pedestrian carrying a Bluetooth beacon.