I read an article by Bob Trebilcock in Modern Materials Handling titled “Automation: Grocery giant Kroger changes the distribution game,”and I was amazed by what Kroger has accomplished, and the risks the company has taken.
Kroger implemented an automation system “that can receive and put-away full pallets, then break them down and rebuild them into store-ready mixed pallets according to how they will be put away on the shelves in a store aisle. It all happens with almost no human intervention: Operators typically touch a product once when lift truck operators unload pallets at the receiving dock and once again when they are loaded at the shipping dock for delivery to Kroger’s stores.”
The system automatically builds efficient mixed-case pallets for individual stores in such a way that product is not damaged, the cube of a truck is maximized, and the pallet gets to the store in an “aisle-aligned” manner based on the store’s planogram. As described in the article, “Aisle-aligned means that cartons are loaded on a pallet in the sequence they’ll be put away on the shelves in a specific aisle, with the top layers at one end, the middle layers in the middle of the aisle, and the bottom layer at the other end. But, the system would also understand constraints like crushability so that it doesn’t stack cans of soup on top of Japanese noodles on the pallet.”
Wow! What a feat of engineering!
But the risks seem very large to me. Often, when making these types of automation investments, companies need to make multi-year assumptions about volume growth and order profiles. For the first few years, many companies find that the automation system is underutilized. Then, when order volumes and profiles reach an optimum level, the system is cost efficient for several years, but as volumes grow past this optimum level, companies need to make difficult decisions. In Kroger’s case, the company must have made assumptions about the number of stores its DC would support. If stores were to close in the coverage area, Kroger may be left with more capacity than it can utilize cost effectively.
What if you could achieve the same results as Kroger with its big automation project, but with a much lower risk profile? What if as volumes grow, you could add new machines just like you add new forklifts and workers in manual warehouses?
Some of these “what ifs” are beginning to emerge. I recently had a conversation with Toby Rush of Rush Tracking Systems. Toby told me of a test project underway in Germany that is evaluating the feasibility of unmanned “robotic” forklifts that can go to the right pick location, move their forks up to the proper height, pick a pallet, and move it to the correct shipping location. A German company is testing this technology at its warehouse during third shift, and using traditional manned forklifts the rest of the day. The unmanned forklifts use optical location sensing and tag readers, proximity and height sensors, and remote control automation. In the longer term, these unmanned forklifts could be AS/RS killers.
One part of the technology stack is optical sensing technology. Both Toby and some of my colleagues here at ARC have sent me video clips about Microsoft’s Project Natal for Xbox 360. As amazed as I am by what Kroger implemented, I’m even more amazed by this technology. What these videos show are kids playing Xbox games without any controllers. The video games are controlled using voice recognition and optical sensors that sense what the players’ hands and/or feet are doing. Watching these videos makes me believe that the days of unmanned forklifts may not be that far away, even if this pilot test in Germany does not succeed.