Ankor is a wholesaler in Leidschendam, The Netherlands that specializes in tools, hardware, and gardening equipment. In its warehouse, it stores over 18,000 products with diverse characteristics. Ankor categorizes these products in three groups: (1) the 384 fastest-moving products, (2) products whose longest edge is over 80 centimeters, called nonconveyables (Nocos), and (3) regular products that do not fall within the other two groups.
Ankor’s warehouse is under continuous pressure to improve its efficiency while it is confronted with several specific requirements in its order picking. Its management realized that order picking is a major contributor to its operational costs, and travel time for order picking time depends largely on operational procedures. Hence, the goal of the research team was to investigate possibilities for improving the operational procedures, in particular combinations of routing and storage policies. The research team mainly focused on the largest group of regular products, which contains 17,000 of the total of 18,000 products.
Ankor believed that the routing policies (Design of route that order pickers need to follow when picking the products at the warehouse) and storage policies (Ways to store the products to enhance the easiness for order-pickers easier to pick the products) will affect the delivery speed of the products and thus the cost of production. Thus, simulation model has been created to calculate the length of the route created by a number of routing policies. And the company can discover the best routing and storage policy.
Ankor considered heuristic routing policies: the mid-point policy, largest-gap policy, S-shaped policy and combined policy.
For the mid-point policy, the picker occasionally travels through an entire aisle or to go either to the next cross aisle or to the end point of the route. Thus, for the ground floor, the mid-point heuristic policy is equivalent to the traditional routing policy plus some shortcuts.
For the largest-gap policy, it is the same as the mid-point policy, except that the picker goes up an aisle as far as the largest gap instead of to the middle. The largest gap is the largest part of the aisle without pick locations for this route.
For the S-shaped policy, it is quite different from the previous two. The order picker travels through a cross aisle to an aisle with a pick location, travels through that aisle to another cross aisle, continues along this cross aisle to the next aisle on the same side with a pick location and travel through this aisle to return to the original cross aisle.
The combined heuristic policy decides for each aisle whether the picker should traverse it entirely or return to the cross aisle from which he or she entered the aisle.
For the storage policy, the company first classified the products into 3 parts, class A contained up to 70% of the total picks and at most 15% of the products, class A and class B together contained up to 90% of the picks and at most 50% of the products, class C contained the remaining products.
The company filled class A first. Repeatedly it added to class A the product with the highest number of picks that was not yet assigned to a class until the number of picks in class A reached 70% of all picks or until 15%of the products were in class A, whichever came first. They filled classes B and C with products consecutively in the same fashion.
Finally, Ankor grouped the routing policies according to the storage policies they required. S-shaped and combined policies were grouped together as the picker only need to enter an aisle once. The storage policy they used is called combined storage policy. And mid-point and largest gap policies used midpoint storage policy because they required a picker to enter an aisle twice.
There are three...