HYBRID BIM–AI FRAMEWORK FOR AUTOMATED CONSTRUCTION SAFETY RISK PREDICTION

Abstract

Construction sites are also a risky place because of the dynamic nature of the work environment, multiple simultaneous processes, and high levels of human-machine interaction. Traditional methods of managing construction safety are mainly reactive and they are based on manual inspection and historical study that does not usually ensure the prevention of accidents before they happen. To overcome these shortcomings, this paper suggested a Hybrid Building Information Modeling-Artificial Intelligence (BIM-AI) system to enable automated safety risk prediction in construction. The suggested framework will combine space and time information obtained by 4D BIM and AI-based predictive models to anticipate and measure risks related to safety in construction processes. BIM allows geometric organization of representation, schedules and workspaces, whereas AI models observe recurring risk patterns based on past occurrences of accidents and other site-specific variables. The estimated risk levels are displayed right within the BIM environment to facilitate the process of making informed and intuitive decisions. Experimental findings also reveal that the suggested BIM-AI framework is much more accurate in predictions, reliable, and efficient in inferences, than the use of rule-based BIM and independent AI solutions. The framework offers a scalable and efficient proactive approach to safety management in contemporary construction works.