A Comprehensive Review of the Literature on Machine Learning-Based Road Safety Prediction Techniques for Internet of Vehicles (IoV)-Enabled Vehicles

Abstract

The goal of this comprehensive assessment of the literature is to ascertain the state-of-the-art in machine learning (ML)-based vehicle safety measure prediction, obstruction estimate, and road traffic analysis the vehicle-related internet (IoV). In particular, we concentrate on confirming the necessity and extent of federated studying in this area. A decentralized machine learning method called federated learning enables numerous edge devices to train a shared model together while storing the data locally. We looked through numerous scholarly databases and a few chosen peer-reviewed papers on the subject. The paper outlines the benefits and future prospects of federated learning in road traffic analysis and vehicle safety while highlighting the current drawbacks and restrictions of the conventional centralized ML systems. In addition, we examined the state of the art studies in federated learning for road traffic analysis at the time and noted any knowledge gaps as well as potential future study areas. The results of this paper show the extent to which federated learning is required in the subject of road traffic analysis and vehicle safety, as well as how it can be used to get around some of the drawbacks of centralized machine learning techniques.