Dual-channel optimization of prefabricated construction supply chains using two-stage stochastic programming

Prefabricated construction supply chains face uncertain daily demand and limited production capacity, but existing models often assume sufficient capacity or deterministic demand. This paper presents a two-stage stochastic programming model with a dual-channel strategy that combines direct factory-to-site delivery with temporary warehouse storage. The first stage determines production rates, fleet sizes, warehouse capacity, and advance-production lead time. The second stage adjusts daily production, inventory, and shipments for each demand scenario. Computational experiments compare the strategy with just-in-time (JIT) delivery and a warehouse-only configuration. The dual-channel strategy lowers expected total system cost relative to the warehouse-only configuration in all tested conditions, by 7.58% on average. Its average saving over JIT is 4.81%, but JIT is more cost-effective with longer construction durations or sufficient production capacity. The dual-channel strategy has an essentially zero expected shortage rate, whereas JIT averages 0.389%. A case study illustrates the model’s project-level application.