Blockchain-Enabled Multi-Objective Construction Supply Chain Optimization
Abstract
Construction procurement decisions inherently trade off cost, transportation and holding expenses, supplier reliability, and delivery lead time a structure well suited to multi-objective optimization, yet rarely modeled with the genuine nonlinear mechanisms (quantity discounts, capacity congestion, reliability degradation under load) that characterize real supplier behavior. This paper formulates a five-objective, mixed-integer nonlinear procurement model for a construction supply chain in Zelten, Libya, incorporating blockchain-enabled sourcing as a per-supplier adoption decision with an explicit adoption cost, genuine multi-period inventory dynamics, and a reliability service-level constraint. We report a practical finding relevant to similarly constrained problems: NSGA-II, despite being standard practice for this problem class, twice failed to maintain solution diversity under tightly binding constraints, collapsing to near-single-point Pareto sets; a weighted-sum sweep using multi-start Sequential Quadratic Programming (fmincon) recovered a genuine, wide Pareto front in both cases. Results show that blockchain-enabled sourcing does not merely shift average performance, but shifts the achievable trade-off frontier toward lower unreliability and shorter lead times: blockchain-enabled solutions reach lower unreliability (minimum 244 vs. 363) and shorter lead times (minimum 7,370 vs. 9,237 supplier-days) than the best baseline solutions, at a higher minimum procurement cost (353,040 vs. 342,360 LYD), even after accounting for adoption costs and the model's added operational realism. The case study combines verified project quantities with documented assumptions and is illustrative.