Keywords = Metamodelling

A Meta-Model based-Simulation Optimization approach for controlling rework delay duration in construction: An Intelligent Metamodeling Approach

Articles in Press, Accepted Manuscript, Available Online from 25 October 2026

https://doi.org/10.22116/jiems.2026.588428.1639

Mohammadreza Davoodabadi Farahani, Alireza Habibi, Hossein Amoozad Khalili, Mohammad Ehsanifar

Abstract Background and Objective: Rework is one of the main contributors to cost overruns, delays, and reduced productivity in construction projects. Since this phenomenon has a nonlinear and stochastic nature, conventional analytical methods do not provide sufficient capability for modeling and controlling it. The objective of this study is to propose an integrated metamodel-based simulation–optimization framework to control the duration of rework-induced delays in construction projects.
Research Method: This study employs a hybrid approach combining mathematical modeling, discrete-event simulation, design of experiments (DOE), central composite design (CCD), regression-based metamodeling, and heuristic optimization. The Modified NBI method is also used to address the multi-objective optimization problem. The research data were collected from a five-story building project in Tehran as well as 55 construction projects in the Middle East during 2015–2020. The results indicate that factors such as rework frequency and duration, workforce experience, conflicts among project stakeholders, payment delays, and political conditions are among the most influential contributors to rework. Moreover, the proposed metamodel transforms the simulation objective function from a black-box representation into a more explicit and solvable form, thereby reducing the computational burden of the optimization process.The findings suggest that integrating simulation, metamodeling, and optimization can provide an effective approach for controlling and reducing rework-related delays in construction projects. This framework improves decision-making accuracy while enabling the identification of optimal and Pareto solutions at a lower computational cost.