A Real-Time Detection and Localization System of Unexpected Events Using Crowd-Sourcing

Through crowd sourcing, we have been developing a realtime detection and localization system of unexpected events such as traffic accidents. For such a study, we previously proposed a crowd-sourced system called Smart and Quick Unexpected-Event Detect (SQUED) system that used smartphone sensor data for the detection and localization of unexpected events. When SQUED system users encounter an event, they point their smartphones toward the direction of the event. SQUED system can detect the location and time of the event using smartphone sensor data. In this study, we further developed our SQUED system to realize real-time detection and localization function. In addition, we developed the pseudo-user generator to evaluate our systems. Our pseudo-user generator can simulate the environment in which the number of users has increased. As a result, it was confirmed that our real-time system in combination with the pseudo-user generator could perform event detection 100 times faster than our previous system.