Risk assessment in the context of modern methods of construction leveraging data analytics

Abstract

As the construction industry moves towards more innovative, sustainable, and faster construction techniques, Modern Methods of Construction (MMC) has been regarded as a solution for meeting these demands. Nonetheless, under immense pressure, decision-makers' opposition caused by oblivious stakeholders makes for tenuous circumstances for innovation. Yet, the literature lacks a comprehensive approach for cost overruns risk assessment of implementing MMC. The study, therefore, aims to encourage the further take-up of offsite MMC in future projects of the housing sector by primarily soliciting opinions from experienced professionals. In achieving this endeavor, the study (1) identified and prioritized risk factors; (2) revealed the underlying categories; (3) proposed ways to prioritize risks; and lastly (4) developed an AI-based risk assessment model. Through the adoption of the generative adversarial networks, an Artificial Neural Networks (ANN) model will also be developed. Study findings revealed the significant dissimilar criticality levels of risk factors. The top seven most risky factors are Safety Hazards, Direct Costs, Poor Understanding, Quality Monitoring, Scheduling and Planning, Site Layout, and Machinery and Technology with an overall frequency of occurrence of 0.736, 0.733, 0.730, 0.725, 0.708, 0.705, and 0.702, respectively. The study brings to light the inadequacy of the current industry and indicates that future research opportunities lie in the adoption of MMC. The study adds value to the literature by exploring and capturing hidden trends and patterns related to conditional dependence between risk factors. The results aid the industry stakeholders to prioritize risk factors to develop risk response measures. Accordingly, decision-makers will be capable to distribute the contingency budget on more uncertain events, which will potentially facilitate achieving project objectives and avoid racking up substantial losses.