Naturalistic driving data for a smart cloud-based abnormal driving detector

This paper examines a possible implementation of video and radar data into smart abnormal driving behavior detecting systems. A prevailing research is SafeDrive, which positions the application in a successful trajectory in identifying driving anomalies to improve societal transportation safety. However, there are limitations of road and environment data that could allow mobile phone-based applications to accurately evaluate driving abnormality. In this paper we propose utilizing external research from the Second Strategic Highway Research Program (SHRP2), which can provide video and radar data for the improvement of detecting and evaluating a driver's atypical behaviors. The collection of data allows a deep study on safety, renewal, capacity, and reliability in a vehicle's actions that could permit transportation organizations to efficiently assess security operatives. Additionally, the evaluation of gathered data encompasses tools that determine and discuss features of both a vehicle and its driver, which influence driving styles linked to the near-crash or crash events. The Naturalistic Driving Data (NDD), supplies detailed information and examination of a human-vehicle-interaction to better comprehend abnormal driving behavior and critical factors that lead to most traffic accidents.