Smart Weather-Aware and Safety-Optimized Route Planning Using Machine Learning and Spatial Risk Analytics

The environmental characteristics and weather conditions of cities directly influence the safety of city streets.Conventional navigation systems are optimized for the quickest possible travel times and shortest distances, without consideration for safety or the risks associated with travel through these areas.In this research, we propose an urban road safety framework that utilizes real-time environmental data and accident-prone locations to provide a comprehensive safety rating for a traveler's route based on the severity of potential accidents.We utilized a random forest model to develop a structured database of both road and environmental characteristics, which has a mean absolute error of 0.96, a root mean squared error of 1.13, and an R² score of 0.828, all measured against a 1-10 severity scale.For determining whether the risk of being involved in a serious accident is categorized as low, medium, or high, our model produced a classification accuracy rate of 78.67%.The risk prediction is then combined with environmental penalties for adverse conditions and penalty values for accident-prone areas to create a safety index at the route level.Our experimental testing was performed on various routes throughout Delhi and demonstrated that the proposed framework produces safety-conscious routes while maintaining little variance in travel time compared to time-optimal routes.These results demonstrate that utilizing predictive risk models with environmental intelligence can improve decision-making in urban navigation to reduce the risk of accidents.