Driving-behavior monitoring using an Unmanned Aircraft System (UAS)

Abstract

Abnormal driving behaviors have been used as cues to identify Driving While Intoxicated (DWI) drivers and prevent DWI-related accidents. Currently law enforcement officials rely on visual observation for detecting such behaviors and identify potentially DWI drivers. This approach however, is subject to human error and limited to vehicles in a very small region. To overcome these limitations, we propose to use an Unmanned Aircraft System (UAS) for driving-behavior monitoring to prevent accidents and promote highway safety. A high-resolution optical camera on the UAS is used to capture the movement of vehicles on the road. The vehicle trajectories are tracked from the captured videos to identify the misbehaviors. This allows for a quicker response time from law enforcement. A risk model is also developed for analyzing the severity of misbehaviors while providing a basis to take appropriate action to reduce dangerous activity. We developed a UAS prototype to showcase the practicality and effectiveness of the proposed system. It was demonstrated that dangerous driving activity on highways can be effectively and timely detected and analyzed.