Optimally loading tow trains for just-in-time supply of mixed-model assembly lines

In today's mixed-model assembly production, there are two recent trends-namely, increasing vertical integration and the proliferation of product variety-that shift focus to an efficient just-in-time part supply. In this context, many automobile manufacturers set up decentralized logistics areas referred to as supermarkets. Here, small tow trains are loaded with parts and travel across the shop floor on specific routes to make frequent small-lot deliveries that are needed by the stations of the line. This article investigates the loading problem of tow trains, which aims at minimizing inventory near the line while avoiding material shortages given the limited capacity of the tow trains. An exact polynomial-time solution procedure is presented and interdependencies with production planning, that is, the sequencing problem of product models launched down the line, are investigated in a comprehensive computational study.