RISK MANAGEMENT IN CONSTRUCTION PROJECTS USING BIM AND ARTIFICIAL INTELLIGENCE: A COMPARATIVE STUDY OF DEVELOPED AND DEVELOPING CONTEXTS

Abstract

The construction sector continues to rank highly among the sectors in the world economy, which are highly risky and poorly digitalized. There is always a persistent threat of cost overrun, time delays, design defects, and safety hazards. BIM and artificial intelligence have become two key pillars of Construction 4.0 as means to take data-driven and proactive as well as predictive actions in risk management. This paper gives an overview and comparison of BIM and AI for construction risk management in both developed and developing countries, focusing particularly on Pakistan. The analysis is based on a systematic review of fifty research papers and reliable reports in peer-reviewed journals from 2018 to 2026 on the subject. The paper makes a comparative assessment of technology adoption, benefits, challenges, and enablers in both matured markets such as the United Kingdom, the US, Germany, and Australia and emerging markets like Pakistan, India, Nigeria, and Southeast Asia. The findings illustrate a clear divergence wherein the adoption of BIM is above 70% in the UK after being mandated by the government in 2016, yet in Pakistan, there is 63% awareness with only 17% usage and use of AI in construction risk management is still at a primitive stage in both contexts, but more advanced in developed countries. The findings illustrate the fact that the constraint in developing countries is mostly institutional, including the lack of any mandate, fragmented standards, training pipeline and awareness, as opposed to technical issues. Based on the findings, the paper suggests an integrated multi-layered BIM-AI risk management framework that could be adapted in rich and scarce resource contexts consisting of Data Integration, AI Analytics Engine, Digital Twin Simulation and Decision Support Layer with continuous feedback mechanism. Insights from case studies reveal that integrated application will result in significant reduction in design mistakes, rework and schedule slippages, while also improving hazard identification