Order picking is a costly warehousing process. Pickers follow routes in the warehouse pushing a trolley, to collect items to form customer orders. To save time, several orders might be batched, i.e., put together to be collected in one single tour. The key decision is therefore how to group customer's orders to minimize the total walked distance.

Two typical examples of warehouse layouts are used for the benchmark datasets you can download below. Figure (a) shows a regular rectangular layout made of three vertical aisles and three horizontal cross-aisles. Products are located on both sides of vertical aisles; cross-aisles enable the order picker to navigate in the warehouse. It is the setup of the Foodmart data set. Figure (b) shows an acyclic layout where pickers are not allowed to backtrack. It is the setup of the HappyChic data set.

You can also download below the .jar of our picking optimizer and the solution files for all instances.