EMOCS-CRO: a hybrid multi-objective optimization framework for sustainable construction planning and execution-mode selection

The construction industry significantly contributes to global carbon emissions, creating pressure to reduce the Embodied Carbon (EC) of buildings without compromising structural integrity or construction-related quality. However, most existing optimization approaches address only one or two objectives, such as cost, energy, or carbon, rarely integrating dataset-derived quality and construction-related quality within a unified multi-objective framework. This study proposes a hybrid metaheuristic framework, Enhanced Multi-Objective Cuckoo Search (MOCS)-tuned Coral Reefs Optimization (EMOCS-CRO), to simultaneously optimize dataset-derived quality and EC reduction in sustainable building design. The framework combines the global exploration capability of MOCS with the adaptive local refinement of Coral Reef Optimization. The Sustainable Architecture Optimization Dataset from Kaggle, comprising 69 office building activities and 29 highway construction activities, was used for evaluation, with the model developed in Python 3.11. EMOCS-CRO was benchmarked against MOSGO and eMOGOA using Diversification Measurement (DM), Mean Ideal Distance (MID), Spread (SP), and Hyper-Volume (HV) metrics. The proposed method achieved higher DM (30.137 for Case 1, 50.137 for Case 2) and HV (0.927 and 0.948), with lower MID (0.655 and 0.555) and SP (0.374 and 0.574), outperforming existing methods in both cases. EMOCS-CRO offers a scalable, practical decision-support tool bridging surrogate structural-performance engineering with sustainable architectural design.HighlightsProposes hybrid EMOCS-CRO for sustainable architectural optimization.Simultaneously optimizes strength, esthetics, and embodied carbon.Achieves higher DM and HV with lower MID and SP values.Enhances carbon reduction in office and highway projects.Outperforms MOSGO and eMOGOA in multi-objective performance.