A Systematic Review of Context-Aware Artificial Intelligence and Machine Learning Approaches for Road Risk Prediction

Abstract

Road traffic accidents continue to be a major contributor to global mortality and economic loss, particularly in developing nations. The increasing availability of real-time data and advancements in Artificial Intelligence (AI) and Machine Learning (ML) have enabled the development of intelligent systems for predicting road risks and analysing driver behaviour. This study presents a comprehensive systematic review of context-aware AI and ML models applied to road risk prediction between 2015 and 2025. A total of 120 peer-reviewed articles were analysed to identify trends, methodologies, datasets, and research gaps.

The review highlights the transition from traditional machine learning models such as Support Vector Machines (SVM) and Random Forest to advanced deep learning architectures including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Transformer-based models, and Graph Neural Networks (GNN). Despite significant progress, several limitations persist, including the lack of multi-modal integration, insufficient contextual awareness, absence of real-time deployment, and limited explainability.

To address these challenges, this study proposes a Context-Aware Intelligent Machine Learning (CAIML) framework that integrates behavioural, environmental, and physiological data. The proposed model demonstrates superior performance with high accuracy and real-time inference capability. The findings provide valuable insights for researchers and practitioners in intelligent transportation systems and highlight future directions for enhancing road safety through AI-driven solutions.