Modeling and solving integrated assembly line balancing, assembly line feeding, and facility sizing problems

The individual research domains of assembly line balancing and feeding have received considerable attention in recent years, furthered by a continuing trend toward mass customization. This research extends the limited literature on the simultaneous consideration of line balancing and feeding while substantially broadening the scope concerning the number of assembly line feeding policies and the incorporation of facility sizing decisions. This large-scale problem is modeled using mathematical programming techniques and solved using a logic-based Benders' decomposition. We used a combination of real-world and re-engineered data from a car manufacturer to conduct numerical experiments. The major findings reveal that integrated decision-making may lead to a substantial cost reduction of up to 20% in this case. Furthermore, the study explores subsets of feeding policies to reduce the amount of different material flows within the factory while considering their associated costs. Our findings reveal a surprising difference in the importance of individual line feeding policies, specifically identifying boxed-supply as a pivotal policy for ensuring feasibility and reducing costs.