PREDICTIVE ACCIDENT PREVENTION USING TELEMATICS AND AI

Abstract

The article is devoted to the study of trends in road accidents and directions of problem prevention through the use of telematics data and artificial intelligence technologies. The purpose of the article is to substantiate the possibilities of using telematics data and artificial intelligence to predict road accidents and improve road safety in the USA. During the scientific study, general scientific methods of cognition were used, in particular, analysis, synthesis, comparison, generalization, systematization and classification. The results of the study show that several key risk groups have been identified in the field of road accidents in the USA. It was studied that the highest mortality rate among men aged 20–24 is 27.9 per 100 thousand population, which indicates the need for separate behavioral monitoring of this category of drivers. It was studied that modern models of road accident prediction cover not only speed or violation of rules, but also risk perception, behavioral instability, workload, distance, relative speed and trust in automation. It is shown that models of deep learning, behavioral entropy, simulation evaluation, 3D modeling, mental load analysis, microscopic state of motion and car-following can already quantitatively assess dangerous driving patterns and predict risky situations. It is generalized that the author's model AFSA Method combines Calibration Phase Algorithm, AI Risk Scoring Engine, Behavioral Feedback Layer, Fleet Risk Intelligence Layer and Empirical Validation Layer - these are modules that are able to provide prediction of accidents and reduce their probability. The practical significance of the author's model lies in the ability to change the situation on the road before the accident. For drivers, this can mean timely warning about dangerous driving style. For fleets, this creates a tool for risk control and prevention. For insurance models, this provides more accurate behavioral analytics. In a broader sense, such an architecture can become an additional tool for reducing accidents, injuries and human losses on US roads.