Abstract
Traditional construction contractor selection often relies on the "lowest-bidder" approach, which frequently overlooks qualitative risks and long-term project health. This paper proposes an integrated multi-criteria decision-making (MCDM) framework utilizing Fuzzy Analytical Hierarchy Process (FAHP) and Fuzzy TOPSIS. By employing linguistic variables and Triangular Fuzzy Numbers (TFNs), the model captures the inherent uncertainty in human judgment. A case study involving three contractors across five criteria—including sustainability and reputation—demonstrates the model's ability to provide an objective, data-driven ranking that prioritizes project quality and risk mitigation over simple cost-minimization.
Introduction: The Pitfalls of the Lowest Bid
In the construction industry, selecting a contractor based solely on the lowest price often leads to a "race to the bottom," resulting in project delays, safety issues, and compromised quality. The primary challenge in moving toward a more holistic selection process is the difficulty of balancing quantitative data (like credit ratings) with qualitative assessments (like reputation or management experience). Traditional decision-making methods often struggle with the subjectivity and ambiguity of expert opinions. This research addresses this gap by applying fuzzy logic to model the "fuzziness" of human intuition, transforming vague linguistic terms into actionable mathematical data.
Methodology: A Two-Stage Fuzzy Approach
The proposed framework integrates two powerful MCDM tools:
- Fuzzy Analytical Hierarchy Process (FAHP): This stage determines the relative importance (weights) of the selection criteria. Experts perform pairwise comparisons using linguistic scales, which are then converted into Triangular Fuzzy Numbers (TFNs). This approach accounts for the overlap and uncertainty in human judgment that "crisp" numbers cannot capture.
- Fuzzy TOPSIS: Once weights are established, the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) ranks the candidates. It identifies the contractor closest to the "Positive Ideal Solution" (maximum benefit across all criteria) and furthest from the "Negative Ideal Solution" (maximum risk).
Modeling Uncertainty with Linguistic Variables
To bridge the gap between human language and mathematical modeling, the study utilizes a scale of linguistic terms. These terms allow experts to express preferences naturally, which the system then maps to specific fuzzy intervals.
Case Study and Data Analysis
The methodology was tested on a comparative analysis of three contractors evaluated against five key criteria: Sustainable Practices, Client Testimonials, Creditworthiness and Rating, References from previous clients, and Project Management experience. Data was collected via structured questionnaires from ten industry experts. The following table illustrates the qualitative input provided for the three candidate contractors, showcasing the variety of subjective ratings the model must process.
Results and Discussion
The integrated model successfully defuzzified the complex expert inputs to produce a clear ranking.
- Contractor 3 emerged as the optimal choice with a Closeness Coefficient (CCi) of 0.6022.
- Contractor 2 followed with a CCi of 0.4268.
- Contractor 1 ranked lowest with a CCi of 0.3163.
A significant finding was the high weight assigned to "Sustainable Practices," which achieved a fuzzy geometric mean of 7.177. This reflects a modern industry shift where environmental responsibility is increasingly viewed as a critical indicator of a contractor's overall competence and long-term viability.
Conclusion
The FAHP-TOPSIS framework provides a robust alternative to traditional selection methods by mathematically formalizing expert intuition. While computationally more intensive, the model offers a "safety buffer" by explicitly calculating the distance from worst-case scenarios. Future research will explore the integration of Artificial Intelligence to further automate weight assignment and scale the model for larger, more complex infrastructure projects.
Summary updated