Improving unit-load warehouses with new aisle designs

Unit-load warehouses have been in use for decades, and they still play a critical role in the supply chain. At the same time, the recent work of Gue and Meller (2006) has changed the way we think of the aisle design problem. By relaxing implicit constraints used for years in warehouse design by both researchers and industry, Gue and Meller have developed innovative new designs that improve upon traditional aisle layouts by reducing expected travel in unit-load warehouses.;In Contribution 1 we develop an analytical expression for the expected travel-between distance in a fishbone warehouse. Combining this expression with the previous expression for single-command travel distance, results in an analytical expression for expected dual-command travel distance. This model of dual-command travel allows us to optimize fishbone aisle designs for dual-command use, and provide two general design guidelines. The best, or nearly-best, design is obtained (1) by choosing a warehouse shape that is approximately a square half-warehouse, and (2) by extending the diagonal cross aisle to the upper corners of the picking space. We also compare the performance of fishbone under dual-command operations to traditional warehouse designs and show reductions in travel distances of 10-15%.;In Contribution 2 we investigate the performance of flying-V and fishbone for both single-command and dual-command operations under turnover-based storage policies. We compare their performance to three traditional aisle layouts under random storage and under turnover-based storage using three ABC curves: 20/40, 20/60 and 20/80. The flying-V cross aisle is preferred over the traditional layouts for single-command travel in most instances, but is generally not recommended for warehouses that perform 100% dual-command operations and use a turnover-based storage policy. The fishbone design performs better than the traditional designs for all the warehouse sizes and operating conditions we evaluated. For single-command travel, the improvement is 10-20% for medium-sized to large warehouses. Improvement in dual-command travel distances is 6-16% across all warehouse sizes. Our study has also shown that for all five aisle layouts, the warehouse shape and aisle structure that performs best under random storage also performs well under turnover-based storage for both single- and dual-command operations.;Contribution 3 investigates a new aisle design for unit-load warehouses that perform 100% dual-command operations. We refer to this design as "fishbone triangle" because it is a fishbone design with an extra cross aisle inserted in the upper region to facilitate travel between storage locations. We find that the inserted cross aisle reduces expected travel-between distances over a fishbone warehouse for medium-sized and large capacity warehouses. Fishbone triangle improves on traditional warehouses with middle cross aisles for those same warehouse sizes by 10.6% and 11.7%, respectively, but requires a larger warehouse area.;The new aisle designs incorporating the flying-V cross aisle and fishbone aisles were developed to reduce travel to a single storage location under a random storage policy. The purpose of this research is to more fully develop the flying-V and fishbone designs by considering issues such as dual-command travel (Contribution 1) and nonuniform storage (Contribution 2). We also investigate a new aisle design to further reduce expected dual-command travel in unit-load warehouses (Contribution 3).