A probabilistic multi-criteria framework for assessing circularity and sustainability in modern methods of construction

Purpose This study aims to provide a comprehensive and objective method for assessing sustainability and circularity in the construction industry, with a specific focus on Modern Methods of Construction (MMC) products. By integrating probabilistic modelling and a hierarchical indicator structure, the research supports informed decision-making under uncertainty. It also explores the interconnectedness of sustainability and circularity dimensions, highlighting how different MMC scenarios perform across economic, environmental, social, and technological criteria to facilitate more balanced and strategic material and design choices. Design/methodology/approach This study develops a Probabilistic Multi-Criteria Decision-Making (P-MCDM) framework that integrates the Analytic Hierarchy Process (AHP), a constrained sampling Monte Carlo simulation, and the MIVES method. The framework addresses uncertainty in weighting and criteria evaluation by generating 1,000 probabilistic combinations of indicators. A four-layer hierarchical model is constructed to assess 12 sustainability criteria, and after reliability analysis, 11 circularity criteria. The approach is applied to three MMC products as three scenarios for comparison. Findings The probabilistic analysis reveals that each MMC product scenario exhibits unique strengths across sustainability and circularity domains. The use of LHS-enhanced Monte Carlo simulations produces a robust distribution of performance outcomes, demonstrating the proposed method’s high reliability and sensitivity. Originality/value This research pioneers a structured, probabilistic framework for MMC products that couples circularity and sustainability. The novel combination of AHP, MIVES, and probabilistic modelling, underpinned by a systematically developed and validated indicator set, offers a scalable and adaptable tool for both researchers and practitioners. Comparing cumulative and probability density functions enables probabilistic rankings and confidence-level insights at multiple criteria levels, supporting more resilient and nuanced evaluation of MMC products under real-world uncertainties.