Date(s) - January 24, 2020
11:45 am - 12:35 pm
406 Weil Hall
Alice Smith, Ph.D., P.E.
Joe W. Forehand/Accenture Distinguished Professor, Industrial & Systems Engineering Department
Abstract: Design of Order Picking Warehouses
This seminar will discuss the formulation and solving of the order picking warehouse design problem. Design variables include the number and placement of both the cross aisles and the pick aisles, the location of the depot, and the aspect ratio of the warehouse for a given storage capacity. For a candidate design, product is allocated to storage location using a new algorithm and then designs are evaluated with a sampling of possible orders by shortest path routing through the candidate warehouse design. The shortest path algorithm has been adapted to reflect human preferences in moving through such a warehouse using a visibility graph. This problem is complex both from an algorithmic perspective and from a computational perspective. We address these challenges through smart order sampling and an efficient meta-heuristic using an innovative encoding and problem decomposition, and also through distributed and parallel computing. Results indicate that many designs currently used in practice are surprisingly effective even though they are simplistic and obvious. However, for certain warehouse environment characteristics non-traditional designs can offer operational advantages. Furthermore, the new travel distance metric of the visibility graph can better reflects actual movement than the traditional aisle center method and can influence the selection of the best design.