Crossdocking is a logistics technique used in the retail and trucking industries to rapidly consolidate shipments from disparate sources and realize economies of scale in outbound transportation. Crossdocking essentially eliminates the inventory-holding function of a warehouse while still allowing it to serve its consolidation and shipping functions. The idea is to transfer incoming shipments directly to outgoing trailers without storing them in between. Shipments typically spend less than 24 hours at the facility, sometimes less than an hour.

Here's how it works: in a traditional warehouse, goods are received from vendors and stored in devices like pallet racks or shelving. When a customer (e.g., the consumer or perhaps a retail outlet) requests an item, workers pick it from the shelves and send it to the destination. In a crossdock, goods arriving from the vendor already have a customer assigned, so workers need only move the shipment from the inbound trailer to an outbound trailer bound for the appropriate destination. The already part should make you think of information system requirements--a chief obstacle to implementing crossdocking successfully.

We have been working on operational and design problems in crossdocking for several years. Our early work focused on terminals in the LTL trucking industry, but our recent work is more oriented toward retail crossdocking. Most of what we do concerns what happens inside the crossdock.

How should managers assign trailers to doors?

There are really two questions here: how should outbound trailers be assigned to doors, and where to put inbound trailers when they arrive?  The first problem is commonly called the layout problem.  In the illustration below, filled squares present receiving, or strip, doors; empty squares represent load doors.  The length of a line represents the relative flow for the destination at that load door.

One way to classify crossdocking operations is according to when the customer is assigned to an individual pallet or product. In pre-distribution crossdocking, the customer is assigned before the shipment leaves the vendor, so it arrives to the crossdock bagged and tagged for transfer. In post-distribution crossdocking, the crossdock itself allocates material to its stores. For example, a crossdock at a Wal-Mart might receive 20 pallets of Tide detergent without labels for individual stores. Workers at the crossdock allocate 3 pallets to Store 23, 5 pallets to Store 14, and so on.

Pre-distribution is definitely more difficult to implement because the vendors of the crossdock must know which customers of the crossdock need what before they send the shipment. This involves quite a bit of information transfer, system integration, and coordination. For a distributor with hundreds of vendors, the problem is very big!

The illustration represents the actual layout of an LTL trucking terminal in Atlanta several years ago.  Because congestion on this dock had been a big problem, managers sought to distribute high-flow destinations to six distinct regions (which they accomplished, obviously).  The problem with this layout is that it led to high costs due to travel--locations with the highest levels of flow were on the ends of the dock.

Below is a layout we constructed with the singular goal of minimizing the average distance between strip doors and load doors.  Notice that high-flow locations are in the center, and strip doors are located both next to, and directly across from, these regions of highest flow.

The problem with this layout is that congestion is sure to be a problem because the highest flow destinations are located in the center of the dock.  (Managers at this crossdock confirmed that this layout was unworkable, due to certain congestion.)

Our research proposed a model that balances the costs of worker travel and congestion.  Below is an example of its application to this crossdock.  Notice that high-flow destinations are more widely distributed on the dock, although they are still “centrally located.”

Using a model such as ours, LTL carriers could expect to reduce labor costs due to travel by about 10%.  (Travel is typically 20-30% of total dock labor costs.)   For more information on this topic, see our papers,

1. ‣John J. Bartholdi III and Kevin R. Gue, Reducing Labor Costs in an LTL Crossdocking Terminal, Operations Research 48 (6), 823-832, 2000.
   
   

2. ‣Kevin R. Gue, The Effects of Trailer Scheduling on the Layout of Freight Terminals, Transportation Science 33 (4), 419-428, 1999.
   

Next page

Sponsored by the Office of Naval Research