Leveraging digital technologies and machine learning for climate resilient buildings : modelling the role of dynamic drivers in mitigating critical barriers

Abstract

Purpose: Digital technologies (DT) and machine learning (ML) offer significant opportunities for the construction industry (CI), particularly in climate resilience (CR) assessment. Despite these potential opportunities, the adoption level of DT and ML remains low in many regions due to the perceived barriers. This study aims to address a critical knowledge gap by investigating the dynamic and dual impact of current drivers on direct adoption and their indirect influence through the mitigation of key barriers towards climate-resilient buildings. Design/methodology/approach: The current study adopted a mixed-method approach. Through an expert survey involving 97 construction practitioners, the partial least squares structural equation modelling method was utilised to explore and validate the proposed model. Findings: Findings confirmed a substantial impact (ß = 0.723, p < 0.000) of the current drivers on the adoption of DT and ML for CR. Additionally, the model demonstrated that the current drivers have a significant positive indirect impact (ß = 0.573, p < 0.000) on adoption by mitigating the critical barriers. This feedback loop suggests the importance of focusing on drivers with dual impact. The analysis indicated that remote sensing applications have the greatest potential for achieving CR. While Building Information Modelling and Digital Twins are significant, their impacts on CR are limited. Interestingly, lack of standardisation is identified as the most critical barrier, as it influences governmental support, which is a primary determinant of adoption. Originality/value: This study makes a crucial contribution to CR objectives in the CI and identifies the existing gaps in Internet of Things integration for achieving CR.