Mixed-level pallets are one of the newest ways distributors and retailers are seeking to reduce replenishment quantities and therefore inventory costs in the supply chain. Building mixed-level pallets economically requires robotic arms that pick cartons from a feed conveyor and place them onto a shipping pallet.
The sequence of cartons arriving to the robot is important for a number of reasons—to keep heavy items under lighter items, for example, or to keep items nearby on the pallet that are nearby on store shelves, which reduces stocking labor in the store.
A mixed-level pallet building robot by Kuka Systems. Notice the feed conveyor in the rear. Cartons must arrive in exactly the right sequence.
Controlling the sequence of arrivals is difficult or very slow if attempted with conventional sortation and conveyor systems. For example, one way to accomplish sequencing is with a loop conveyor. Items enter the loop in any order, circulate for awhile, then are peeled off when “their numbers” are called. This particular solution eats up lots of valuable floor space.
GridSequence is a new way to sequence cartons based on the puzzle-architecture of GridFlow, but with some important differences. GridSequence takes a random stream of cartons, totes, or other unit-loads (the concept is scalable) and assigns each arrival to a buffer position in the grid. As columns in the grid are filled, they depart in sequence. Intra-column sorting is accomplished by puzzle-like movement. The best way to explain it is to show it:
GridSequence has very high density, so doesn’t eat up much floor space. It could even be deployed in multiple levels to increase space utilization. As with our other puzzle-based material handling systems, GridSequence features decentralized control—each conveyor module detects its local environment (occupied? occupied with correct item? etc.), communicates with its neighbors, and negotiates to take appropriate action. The global behavior you see in the video is an outcome of these local decisions.
We are also thinking about other applications of these ideas. For example, in automobile manufacturing, the sequencing of parts, subassemblies, and chassis is a significant challenge that could be met by a system such as GridSequence.