Contractor Selection Criteria for Building Refurbishment Projects through Genetic Algorithm (GA)

The objective of this study is to determine the most ideal composition of criteria to be considered by project owners in selecting specialized contractors for architectural and MEP (Mechanical, Electrical, and Plumbing) renovation works in existing buildings. This research also aims to explore the structure and stages of implementing the Genetic Algorithm (GA) method in the context of optimizing contractor selection criteria, specifically for architectural and MEP repair works. The study was conducted through GA-based analysis utilizing both secondary data (literature review) and primary data (respondent questionnaires). The results of GA for evaluating Architectural criteria indicate that Work Quality and Project Comprehension emerged as the dominant criteria from both the historical (secondary) perspective and the viewpoint of field stakeholders (primary). Meanwhile, the GA result for MEP criteria evaluation revealed that Work Quality and Personnel Competencies were consistently identified as key criteria across both data perspectives. The application of the Genetic Algorithm (GA) in this study has proven effective in evaluating and determining the optimal weights for each contractor selection criterion in building refurbishment projects. The GA parameters used in this study included an initial population of 150 individuals, 5000 generations, a crossover rate of 0.5 using the arithmetic crossover method, and an adaptive mutation rate set at 0.8 in the early generations, 0.5 in the mid generations, and 0.05 in the final generations—using random mutation. The selection process employed the Stochastic Universal Sampling (SUS) method.