Abstract
Effective asset management is crucial for the smooth operation of hospital facilities, particularly in developing countries like South Africa, where infrastructure challenges, resource limitations, and outdated maintenance practices hinder optimal performance. Previous research on various modelling approaches for hospital infrastructure management has yielded mixed results, often lacking contextual relevance and practical integration strategies. There remains a significant gap in frameworks that enable the reliable and effective integration of Artificial Intelligence (AI) into healthcare infrastructure management to enhance operational efficiency and bridge the knowledge gap in the South African built environment industry. Previous research on various modelling approaches for hospital infrastructure management has yielded mixed results, often lacking contextual relevance and practical integration strategies. In particular, there remains a significant gap in frameworks that enable AI's reliable and effective integration into healthcare infrastructure management to enhance operational efficiency and bridge the existing knowledge gap in the South African built environment industry. This study addresses this gap by developing and validating a context-specific AI-driven Asset Information Management (AIM) model for hospital facilities. The objectives include assessing the current status of AIM and AI usage in hospitals, identifying key factors influencing AI adoption, and constructing a model that supports predictive maintenance, asset lifecycle optimisation, and data governance.
A systematic review was conducted on Scopus for bibliometric and methodology analysis. One hundred twenty-four (124) publications were identified from 2004 to 2024. The bibliometric analysis highlights publication and citation patterns, country distributions, and keyword clusters found within these publications. The methodology analysis examined the methods and analytical techniques used in the 124 publications. The theoretical and conceptual framework synthesized existing knowledge, establishing a strong foundation that informs the development of the research frameworkThis was followed by the development of a theoretical and conceptual framework informed by established models such as Technogy Acceptance Model (TAM), Unified Theory of Acceptance and Use of Technology (UTAUT), and Technology-Organization-Environment (TOE).
vi
A quantitative approach distributed a closed-ended MS Forms survey to built environment professionals actively involved in hospital and digitalised facility management. Using judgmental sampling, the survey targeted professionals affiliated with built environment professional bodies. A total of 211 responses were received, providing a substantial dataset for analysis. The survey data underwent analysis using Partial Least Squares-Structural Equation Modelling (PLS-SEM), a method selected to evaluate and confirm the connections between essential conceptual variables, including organisational readiness, management commitment, technical expertise, and the influence of AI on operational effectiveness. The PLS-SEM analysis was performed using SmartPLS 4, ensuring rigorous model testing and hypothesis validation. Additionally, the descriptive statistics from the survey responses were analysed using SPSS version 29, allowing for an understanding of the dataset's demographic characteristics, central tendencies, and basic trends. The descriptive results analysis revealed a low and moderate level of AI tool and system usage in hospital facilities. The results confirmed the significance of 11 of the 17 hypotheses, validating the model’s constructs, including AI capabilities, process automation, data and security, and organisational readiness. These findings demonstrate that AI can significantly enhance hospital AIM systems by enabling predictive maintenance, reducing equipment downtime, improving asset lifecycle management, and strengthening data governance.
The study confirms that using AI for predictive maintenance and automating processes can decrease equipment downtime, leading to cost savings and better facilities provision. The implications of the findings are crucial for hospital facility stakeholders, as successfully addressing key factors such as technical expertise, high implementation costs, and data security concerns can pave the way for the effective implementation of AI in hospital facilities. Organisational commitment and well-structured systems are also necessary for fostering AI adoption. Future studies should explore strategies such as developing cost-effective AI solutions, implementing comprehensive training programs to build technical expertise, and fostering a culture of innovation to overcome the identified challenges. Additional research is necessary to assess the lasting effects of AI on hospital asset management and explore how AI systems can scale in different hospital environments, especially those with limited resources. This study contributes to both theory and practice by offering a validated, context-specific framework for AI adoption in hospital infrastructure management. It provides actionable insights for hospital administrators, policymakers, and built environment professionals, including implementation pathways,
vii
performance indicators, and strategies for overcoming barriers such as technical expertise gaps and high deployment costs. The model’s adaptability across different hospital types and its alignment with operational realities underscore its potential for scalable application in resource-constrained environments. The study concludes that AI can significantly improve hospital asset management by enabling predictive maintenance, reducing equipment downtime, and improving decision-making. Key enablers include technical expertise, management commitment, and robust data security. The validated model offers a scalable, adaptable solution for resource-constrained environments. This study recommends the inclusion of cost-effective AI tools, implementing training programs to build technical capacity, and fostering a culture of innovation. Future research should explore the long-term impacts and scalability of AI systems across diverse hospital settings