AI-Based Accident Severity Detection Scheme for V2X Communication Beyond 5G Networks

Increasing population growth demands increased usage of vehicles making road accidents rise every day. Despite the development of traffic safety measures, the mortality rate with road accidents is prominent. There is a need for an autonomous and intelligent environment that is sufficiently intelligent to analyze a difficult situation and provide solutions to avoid any hazardous incident, which is provided by Vehicle-to-Everything (V2X) communication. One of the significant use cases is predicting the severity level and providing emergency services to the impacted entities in the accidents. Emergency assistance should be provided in a minimal amount of time during an accident scenario. Hence, efficient accident management procedures should be in place for public safety. Many factors comply with traffic accidents and determining the crucial ones determines the severity of the accidents. We propose an artificial intelligence (AI)-based accident severity detection scheme for V2X communication beyond fifth-generation (5G) networks. In our proposed work, a machine learning-based random forest classifier (RFC) identifies influential factors affecting accident severity and predicts the accident severity levels based on classification achieving approx. 80% accuracy. Moreover, after prediction, depending on the accident severity level, closest emergency service is alerted and dispatched to accident location. We integrate the sixth-generation (6G) communication network with the proposed scheme to construct a scalable and latency-resilient solution.