This study addresses an integrated multi-line flexible flow-shop production scheduling and multi-trip vehicle routing problem (MLFFSPS–MTVR) in perishable supply chains, where due date compliance and minimum life-on-receipt (MLOR) constraints are critical. We formulate a new mixed-integer linear programming (MILP) model to minimize total operational costs, including inventory holding, earliness/lateness penalties, delivery, and fixed fleet-usage costs. To overcome scalability challenges, we propose a hybrid genetic algorithm with local search (HGALS) that combines strategic mutations, problem-specific local searches, adaptive delay and departure-time adjustments, and diversity management. Comparative experiments on benchmark instances show that HGALS consistently outperforms three strong metaheuristic baselines, the hybrid genetic algorithm (HGA) of Vidal, the integrated genetic algorithm (GA) of Ullrich, and a multi-path simulated annealing (MPSA) scheme, achieving average cost reductions of 6.5% and 5.9% over Vidal’s HGA, 17.8% and 15.6% over Ullrich’s GA, and 7.9% and 12.6% over MPSA for medium and large instances, with absolute savings up to 89,000 units. A case study inspired by the dairy sector further indicates that heterogeneous production resources outperform homogeneous configurations, load-consolidation strategies reduce costs despite increasing the number of trips, and MLOR enforcement eliminates violations while supporting Just-in-Time (JIT) and Make-to-Order (MTO) strategies. Overall, the proposed framework provides a robust and scalable method for synchronizing production and distribution under MLOR and due-date constraints in perishable supply chains, with potential applicability to broader logistics contexts. • Studied an integrated MLFFSPS-MTVR framework with due date and MLOR constraints. • Proposed an MILP minimizing holding, delivery, earliness, and fleet usage costs. • Developed a hybrid GA with strategic mutation, local search, and diversity control. • Benchmark results show efficiency and robustness over baseline algorithms. • A dairy case study reveals practical insights under varied configurations.
Kaviyani-Charati et al. (Mon,) studied this question.