Key points are not available for this paper at this time.
Construction projects increasingly face interacting risk drivers arising from performance pressure, safety-conditions, environmental exposure, and operational stress. Despite advances in digital monitoring and analytics, project risk prioritisation in practice remains dominated by cost and schedule indicators, often leading to delayed or misdirected interventions. This study develops and demonstrates a governance-oriented framework for construction project risk prioritisation under uncertainty, reframing risk prioritisation as a decision-structuring problem rather than a purely predictive task. A hierarchical, explainable, and uncertainty-aware risk governance framework is proposed. Latent risk signals are extracted from performance, safety, environmental, and operational indicators using unsupervised learning techniques, followed by domain-specific risk modelling. Domain-level risk beliefs are integrated through evidential reasoning to derive transparent project-level risk priorities. The framework is empirically evaluated using the BIM–AI Integrated Construction Project Dataset comprising 1,000 construction projects with indicators related to cost performance, scheduling, structural health monitoring, environmental conditions, and operational resource utilisation. Results reveal that safety-related risk beliefs frequently dominate overall project vulnerability and may precede observable cost-schedule deviations. These findings suggest that governance dashboards should incorporate leading safety and operational stress indicators alongside traditional performance metrics to support earlier intervention. Future research should examine longitudinal applications across diverse project contexts.
Admane et al. (Fri,) studied this question.