Network-Level Safety Metrics for Overall Traffic Safety Assessment: A Case Study

Driving safety analysis has recently experienced unprecedented improvements thanks to technological advances in precise positioning sensors, artificial intelligence (AI)-based features, autonomous driving systems, connected vehicles, high-throughput computing, and edge computing servers. Particularly, deep learning (DL) methods empowered volume video processing extract safety-related features from massive videos captured by roadside units (RSU). Safety metrics are commonly used measures investigate crashes near-conflict events. However, these provide limited insight into the overall network-level traffic management. On other hand, some assessment efforts devoted crash reports identifying spatial temporal patterns of that correlate with road geometry, volume, weather conditions. This approach relies merely on ignores rich information can help identify role violations crashes. To bridge two perspectives, we define a new set (NSM) assess profile flow imagery taken RSU cameras. Our suggests NSMs show significant statistical associations rates. is different than simply generalizing results individual analyses, since all vehicles contribute calculating NSMs, not only ones involved incidents. perspective considers as complex dynamic system where actions nodes propagate through network influence risk for nodes. The carried out using six cameras state Arizona along 5-year report obtained Department Transportation (ADOT). confirm modulate baseline probability. Therefore, online monitoring be management teams AI-based systems control.