A multi-objective possibilistic programming model integrated with a fuzzy inference system and TOPSIS for order allocation based on the consideration of purchasing decision criteria and SDG17

This research presents an integrated decision-making framework designed to enhance supplier evaluation and order allocation under uncertainty within the electronics manufacturing industry. A total of five evaluation criteria were used: cost, quality, delivery, service, and SDG17, which was newly introduced to incorporate environmental and partnership considerations. These criteria and their sub-criteria were validated through a structured survey of 74 procurement and supply chain professionals, ensuring that both traditional and sustainability dimensions were theoretically and practically justified. Supplier performance was evaluated using a Fuzzy Inference System (FIS), which translated expert judgment into quantifiable scores through Mamdani Type 1 IF–THEN rules across 27 rule sets per criterion. The fuzzified outputs were then defuzzified and ranked using the TOPSIS method. Among the five suppliers analyzed in the case study, supplier SB consistently achieved the highest ranking, followed by SC, SE, SD, and SA. These rankings served as the foundation for proportional order allocation planning. To address uncertainties in lead time, quality, and cost, the study employed a Multi-Objective Possibilistic Programming (MOPP) model with lexicographic optimization. Three prioritized objectives: minimizing lead time, maximizing quality, and minimizing cost, were optimized based on survey-derived weights. The model was tested under three key scenarios: optimistic (α = 0), likelihood (α = 1), and pessimistic (α = 0), each reflecting varying degrees of confidence in uncertain data. The results demonstrated distinct trade-offs: the optimistic scenario yielded the shortest lead time of 10.6 days and the lowest cost of $7.96 million, while the pessimistic scenario was more conservative but ensured robustness in volatile environments. A sensitivity analysis was also conducted across α-cut levels from 0.1 to 0.9, at α = 0.7, the order allocation results revealed that supplier SB received the highest share due to its strong performance across cost, quality, and lead time criteria. The proposed formulated model achieved a 27% reduction in lead time of 6.73 days, a 12% improvement in quality, and a 4% cost reduction of approximately $0.63 million. Among all tested values, α = 0.7 was identified as the most appropriate level for this research, providing an optimal balance between model robustness and decision confidence under procurement uncertainty. In summary, the proposed formulated model offers a comprehensive and adaptive approach for supplier selection and material allocation, supporting both operational excellence and sustainable development objectives. The incorporation of SDG17 as a core evaluation dimension reflects the growing importance of sustainability in procurement strategy, making this model highly relevant for decision-makers in modern, uncertainty-prone supply chains.