Intelligent Modeling of the Circular Economy Sustainability Index in Construction Projects Using an Adaptive Neuro-Fuzzy Inference System (ANFIS

Intelligent Modeling of the Circular Economy Sustainability Index in Construction Projects Using an Adaptive Neuro-Fuzzy Inference System (ANFIS

Mohammad Hossein Jafari1 Mohammad Reza Amiri Shahmirani (Corresponding Author)2 Hassan Javanshir3 Hamidreza Rabieifar4

1) Department of Civil Engineering, Construction Engineering and Management, ST.C, Islamic Azad University, Tehran, Iran
2) Department of Civil Engineering, Construction Engineering and Management, ST.C, Islamic Azad University, Tehran, Iran
3) Department of Industrial Engineering, SR.C., Islamic Azad University, Tehran, Iran.
4) Department of Civil Engineering, Construction Engineering and Management, ST.C, Islamic Azad University, Tehran, Iran

IOI Code : AAHF-ACZEEZ
Publication : National Conference on Architecture and Art Technologies, Smart City, Soft and Creative Industries and Technological Culture(icact.ir)
Abstract :
    The circular economy, as a contemporary approach to sustainable development, plays a significant role in improving resource efficiency, reducing waste, and enhancing the environmental performance of construction projects. However, the complex relationships among its influencing indicators necessitate the application of intelligent modeling techniques. Accordingly, this study aimed to model and predict the Circular Economy Sustainability Index in construction projects using an Adaptive Neuro-Fuzzy Inference System (ANFIS) integrated with the Fuzzy Analytic Hierarchy Process (FAHP). This applied research adopted a descriptive–analytical methodology with a mixed-methods approach. Data were collected through questionnaires completed by academic and industry experts. After identifying the key indicators, their relative importance was determined using the FAHP method, while the questionnaire-based indicator data were used to construct the ANFIS model for capturing nonlinear relationships and predicting the sustainability index. A total of 80 questionnaires were distributed, of which 68 valid responses were included in the final analysis. The 68 expert-response observations were randomly divided into 48 training, 10 validation, and 10 testing cases, corresponding to approximately 70%/15%/15% (70.6%/14.7%/14.7%). The findings identified Resource Efficiency (0.2659), Design for Disassembly (0.2304), and Construction Waste Management (0.1744) as the three highest-weighted circular economy criteria after normalization of the six FAHP weights to a total of 1.0000. Within the 10-observation test subset, the ANFIS model demonstrated satisfactory predictive performance, achieving RMSE = 0.039, MAPE = 4.2%, and R² = 0.89. These results support the internal predictive performance of the expert-assessed model but do not establish project-level generalizability. The integrated FAHP–ANFIS framework can therefore support evaluation of an expert-assessed Circular Economy Sustainability Index and may inform construction-related decision-making, subject to validation using independent project-level data.

Keywords : circular economy construction projects adaptive neuro fuzzy inference system (anfis) sustainability artificial intelligence