This study aims to investigate the key risks and barriers associated with the adoption of artificial intelligence (AI) for risk management in modular construction in the Nigerian construction industry. It addresses the need for data-driven risk management while accounting for diffusion-related constraints affecting AI adoption in developing country contexts.
A quantitative research approach was adopted using a descriptive survey design. Data were collected through a structured questionnaire administered to 309 construction professionals in Nigeria. The Relative Importance Index (RII) was used to rank critical risks in modular construction. Exploratory factor analysis (EFA) was used to categorise barriers to AI adoption, while fuzzy synthetic evaluation (FSE) was applied to assess the relative criticality of the identified barrier clusters. Diffusion of Innovation (DOI) theory was conceptually operationalised through a DOI–barrier attribution matrix.
The results indicate that technological risks (RII = 0.91) represent the most critical challenge in modular construction, followed by financial (RII = 0.87), health-related (RII = 0.86), market and investment (RII = 0.84) and project schedule risks (RII = 0.83). EFA identified five barrier clusters explaining 72.55% of the variance. FSE results reveal that Human Resource and Value Concerns constitute the most critical barrier cluster (index = 4.366, coefficient = 0.203), followed by Integration and Collaboration (4.299), Technical and Structural (4.287), Policy Trust and Implementation (4.286) and Awareness and Legal Framework (4.231). The DOI–barrier attribution matrix confirms that complexity and compatibility-related barriers are the primary constraints on diffusion.
This study emphasises the important role of AI-driven risk management in addressing key barriers to modular construction, especially in developing economies. The findings highlight how AI can mitigate financial, safety and technological risks, thus enhancing project coordination, worker safety and supply chain optimisation. Moreover, the study underscores the need for workforce development, particularly in AI training and inter-industry collaboration, to bridge the skills gap and facilitate successful AI integration. This study contributes to the literature by demonstrating that the successful adoption of AI in modular construction requires a holistic approach that integrates technical advancements, policy reforms and stakeholder education to overcome barriers effectively. The primary limitation of this study lies in its focus on Nigeria, which may limit the generalisability of the findings to other developing economies or technologically advanced countries. Given that AI adoption in modular construction may vary across regions with different levels of technological infrastructure, the barriers identified in this study may not be directly applicable to other contexts. Furthermore, as the construction industry and AI technologies evolve rapidly, future studies should examine how emerging AI tools may affect the challenges and opportunities identified in this research.
The DOI–barrier attribution matrix is the primary original contribution, the first to map empirically derived AI adoption barriers to DOI constructs in a modular framework. Combined with the multi-analytical RII-EFA-FSE framework, this study offers a replicable, diffusion-oriented tool to accelerate AI adoption in developing-country construction contexts.
