My office is organized in a haphazard but personally effective way. I’ve attempted to reorganize it from time to time, but inevitably laziness—and disorder—sets in. Still, my personal chaos works for me—I can find what I’m looking for because my mind forms some vague but effective map, based on context, use and history. In fact, the books and papers I use most frequently in a given period end up closer to my desk almost by natural osmosis.
Several years ago, I began to wonder if this principle of organized chaos could apply to warehouses, eliminating the need for laborious organization and maintenance. My then-student Stephen Ho and I asked: Why are warehouses organized? There are a couple of answers. First, to make it easier to find the stuff we need. Second, it makes sense to store large shipments of like goods together.
|
 |
But recent changes challenge these premises. With the growth of online retailers, warehouses receive large shipments of a variety of goods but send out orders that could be single items or a mix of products. This eliminates the need to keep everything together, assuming you can locate the items. Here’s where the second change enters the picture: With RFID, it should be possible to find goods with much less effort than before.
In Stephen’s doctoral thesis, “Fragmented Warehousing,” he assumes that if we can keep track of the products, it makes sense, under certain circumstances, to intentionally fragment, or disorganize, a warehouse. A fragmented book warehouse, for example, would work as follows: When a shipment of books arrives, it’s broken up and the books are placed into any available cubbyholes in the aisles of the warehouse. When an order comes in for, say, two different books, the picker consults a handheld RFID reader, linked to an inventory tracking system, that displays a route minimizing the walking time to acquire both books.
Why does fragmentation help? With copies of each book dispersed throughout the warehouse, there’s a high probability that somewhere in the warehouse, the books in the order will be in close proximity. Furthermore, the fragmented warehouse is more compact, since it can use all available shelf space.
Christian Floerkemeier, a researcher at the MIT Auto-ID Lab, and Isaac Ehrenberg, a grad student, have tackled the other piece that’s critical to the fragmented warehouse: using RFID to track chaotically stored items. We outfitted a custom mobile robot with high-frequency RFID interrogators and localization algorithms, and tested it at a library where all the books are RFID-tagged. We were able to map not only the books on the shelves, but also their order on each shelf.
We are developing an RFID-enabled robot with simultaneous localization and mapping capabilities to enable warehouse managers to know where everything is at all times. The robot could essentially be dropped into a warehouse without ever being told about its surroundings and be able to map its contents. In the future, similar principles could be applied to enable employees equipped with RFID readers to continuously map a warehouse or yard as they go about their daily work.
Sanjay Sarma is an associate professor in the mechanical engineering department at the Massachusetts Institute of Technology and a cofounder of the Auto-ID Center at MIT.