Methodology for multicriteria evaluation of the effectiveness of construction control and supervision: development of indicators of quality, safety, schedule, and cost accounting for uncertainty factors

Abstract

The study proposes a multicriteria methodology for evaluating the effectiveness of construction control and supervision that overcomes the limitations of single-criterion approaches and enables quantitative attribution of the contribution of control measures to achieving project targets for quality, safety, schedule, and cost under external uncertainty. The empirical base comprises 80 residential, commercial, and industrial projects (2017-2023) with analysis of design-and-estimate documentation, work logs, laboratory protocols, and supervisory directives, supplemented by a survey of 65 experts; on this basis, a four-level hierarchical model was constructed with indicator normalization and weights calculated by Saaty’s Analytic Hierarchy Process. Methodologically, the following were implemented: decomposition into the criteria «quality», «safety», «schedule» and «cost» with systems of partial indicators, linear scaling to [0;1], AHP weighting, treatment of uncertainty via fuzzy logic and Monte Carlo for key risks (logistics, geology/weather, regulation, design errors), and aggregation into an integral indicator of the effectiveness of construction control (IIEC). The results show a nonlinear return to control: moving from low to medium intensity reduces above-estimate deviations from 18,74% to 7,21%, schedule slippages from 29,15% to 11,43%, critical defects from 4,88 to 1,92 units per 1,000 m², and LTIFR from 2,13 to 0,87, whereas further intensification yields diminishing marginal benefits. An investment saturation point is identified: increasing the share of control costs above 2,1% of construction and installation works scarcely raises the integral quality (I_q), while the optimum is robustly achieved in the 1,6-2,1% range with a concurrent reduction in outcome dispersion. Modeling shows that rare but «severe» risks (changes in requirements and regulation; unforeseen geology) make the main contribution to delays (up to 110,5 days, 95th percentile), necessitating a shift toward risk-based planning and proactive management of the external environment. Validation on 15 independent projects confirmed the model’s predictive accuracy (mean absolute deviation 3,91% with no systematic bias), substantiating the use of the IIEC as a tool for scenario-based comparison of control strategies and optimization of resource allocation. Practical implications include targeting control to the project’s risk profile, linkage with BIM platforms for automated monitoring, and calibration of supervision intensity in view of diminishing marginal effectiveness.