Paper Title: A survey on intelligent transportation systems for accident detection and risk prediction using AI
Authors: Vivek Sharma
Corresponding Author: Vivek Sharma (viveksmits@gmail.com)/India
Abstract
The complexity of today’s transportation networks has made road safety, accident detection, and risk management increasingly challenging. Intelligent Transportation Systems (ITS) offer a technological base to enhance transportation safety by incorporating sensors, communications, data processing, and intelligent decision-making. This review discusses the application of Artificial Intelligence (AI) for accident detection and risk prediction in ITS by examining information sources, methodologies, accident investigation, and associated challenges. It examines the use of machine learning, deep learning, computer vision, spatiotemporal modeling, and multimodal AI for accident detection, risk prediction, severity analysis, and active safety management. This review of studies suggests that collecting traffic, vehicle, road, environmental, geographical, historical, and driver-related information can improve understanding of collisions and support timely safety decisions. The review also highlights the importance of explainable and reliable AI that provides transparency, dependability, and accountability. However, the obstacles include data heterogeneity, model generalization, dynamic conditions, scalability, quick implementation, privacy considerations, security requirements, and compatible information systems. In conclusion, incorporating AI in ITS may have the power to shift road-safety management from reactive measures to data-driven approaches.
