Large-scale order batching in parallel-aisle picking systems

This paper discusses an order batching formulation and heuristic solution procedure suitable for a largescale order picking situation in parallel-aisle picking systems. Order batching can decrease the total travel distance of pickers not only through reducing the number of trips, but also by shortening the length of each trip. In practice, some order picking systems retrieve 500~2000 orders per hour and include ten or more aisles. The proposed heuristic produces near-optimal solutions with run times of roughly 70 seconds in a ten-aisle system. The quality of the solutions is demonstrated by comparing with a lower bound developed as a linear programming relaxation of the batching formulation developed in this paper. A simulation study indicates that the proposed heuristic outperforms existing methods described in the literature or used in practice. In addition, the resulting order-picking operations are relatively robust to picker blocking. [Supplementary materials are available for this article. Go to the publisher’s online edition of IIE Transactions for the following supplemental resources: (i) formulation of basic RPP from RSB in Section 4; (ii) a comparison algorithm and additional computational results in Section 6; and (iii) a simulation study to quantify the effect on picker blocking in Section 7.]