BARRIERS TO VIRTUAL REALITY ADOPTION IN ENGINEERING SAFETY TRAINING: A DELPHI CONSENSUS STUDY

Virtual Reality (VR) supported by Artificial Intelligence (AI) offers immersive, data-driven opportunities to improve safety training in high-risk engineering environments. Yet, large-scale corporate deployment remains sporadic. This study applies a three-round Delphi process with 15 academic and industry experts from the University of New South Wales (UNSW), China Railway 25th Bureau, and Zhongyifeng Construction (Suzhou 2nd Bureau) to prioritise the obstacles that prevent VR adoption. Eighteen barriers extracted from Round 1 were rated in Rounds 2 3. Consensus, assessed with Kendall’s coefficient of concordance (W), rose from 0.32 to 0.52, indicating strong convergence. High initial cost, inadequate IT infrastructure, limited management support, scarcity of domain-specific VR content, and lack of integration standards emerged as the top five critical barriers. Secondary constraints included workforce resistance, cybersickness, trainer preparedness, and technical compatibility issues. The paper offers evidence-based recommendations, including financial modelling, staged infrastructure upgrades, executive engagement strategies, content sharing consortia, and standard setting initiatives to accelerate safe, scalable deployment of AI-enhanced VR training in engineering organisations.