Stow & pick: Optimizing combined stowing and picking tours in scattered storage warehouses
研究了分散存储仓库中工人同时进行存储和拣选操作的联合路线优化问题，提出了九种策略和精确分支定界算法，相比传统分离操作可节省11.8%至42.0%的行走距离。
To streamline order fulfillment, many e-commerce retailers apply scattered storage in their picker-to-parts warehouses. Instead of putting homogeneous unit loads into racks, they break up the pallets and store individual products within mixed shelves. This increases the probability that (somewhere in the vast warehouses) products, ending up jointly on hardly predictable pick lists, can be picked from neighboring shelves. Scattered storage promises a reduction of picker travel but comes at the price of additional stowing effort. Instead of applying separate workforces, stowing and picking can also be combined. This paper provides an exact routing algorithm for a worker with a picking cart who departs from the depot with multiple bins filled with products to stow, switches to picking underway, and returns bins full of picked products to the depot. We identify nine different combined stowing and picking policies. We provide an exact solution procedure based on the branch&bound paradigm for all these policies, based on the parallel-aisle structure of warehouses. If a constant limits the number of relevant regions (i.e., picking aisle subsections between adjacent cross aisles with a storage position to be visited), this algorithm guarantees polynomial runtime. We apply our algorithm to benchmark the gains of the nine combined stowing and picking policies compared to the traditional approach, where stowing and picking are executed by separate workforces. Depending on the selected policy, average gains between 11.8 and 42.0 % arise. • We investigate the combined picking and stowing in scattered storage warehouses. • We differentiate 9 variants of the combined pickup and delivery routing problem. • To solve the problem, we exploit the parallel aisle structure of warehouses. • We present an exact branch-and-bound procedure with polynomial runtime.