Robotic piece picking takes another step forward ...
During ProMat, I wrote half in jest that I had seen the future of materials handling and it looked like a lift truck. In that article, I was referring to the new vehicles on display at Dematic’s and Egemin’s booths.
Schaefer was showing off a robot picking small items like toothpaste from a conveyor to a tote and from a tote to a tote. Rob Schmit, a Schaefer vice president, told me the system has the mechanical ability to make 2,500 picks an hour and can realistically pick 2,100 to 2,200 hundred items on a sustained basis. That’s a lot of pieces.
I’m fascinated by robotic picking. Yes, there’s a cool factor there, but I also think that the next few years bode well for the automated materials handling industry. I’m not an engineer or a finance guy, but I talk to a lot of end users. I sense in them a greater willingness to look at automation than in the past. That doesn’t mean their boards will pull the trigger and write a check. But I think labor issues are looming larger than ever.
Companies that downsized have found they can get the job done with fewer employees. Add to that the well-known, but less talked about, hassles of retaining people willing to work in a warehouse and you have a situation where companies would prefer to grow their businesses with automation than with more employees if they can make the numbers work.
Enter robotics. Companies like Schaefer, Witron, Kuka and Axium have been implementing solutions to build store-ready mixed SKU pallets, especially in the food and beverage industry. They’ve got that piece of the puzzle down. Piece picking is the next frontier.
Schaefer believes it has solved the problem –the system I saw at ProMat, for instance, has been installed in a piece picking operation in Brazil.
What does it take to do robotic piece picking? According to Schmit there are at least five components to the system. You’ve got the robot. You’ve got the gripper to pick up the item. You have some type of automated storage and delivery system to deliver the individual pieces to the robotic pick station. You have software that is going to synchronize the demand – the items needed to fill an order – with the delivery system. And, finally you have the vision system that looks at a group of products on a conveyor or in a tote and determines which of those items is going to be picked up and put to a shipping container.
Four of those five are old hat. Robotic case picking solutions are pretty good at getting the right item from an automatic storage system onto a pallet in the right sequence for an order. In piece picking, the trick is the vision system. “Most vision systems have been associated with quality issues,” Schmit says, “like spotting a crack or a flaw on a manufactured item. Our application can find the right product, orient it the right way and then grab it.” Schaefer, he adds, has a team dedicated to vision systems that tie directly to the robot. “You can’t have one without the other,” he says. “Vision is a core competency for us.”
The semi-automated equivalent of a robotic piece picking solution is goods-to-person piece picking. Those applications are usually associated with slow-moving items. Meanwhile, A Frames are often associated with fast-moving items. Where does robotic piece picking fit. “It works best if you bring all of your piece picking operations together into one solution,” Schmit says. “If you want automation for only your slow movers, goods-to-person has a better return. If you’re only interested in fast-moving parts, A frames do a good job.” The ROI in robotic piece picking, on the other hand, comes in part from the fact that you only have to install and maintain one automated solution. “If I put in a robot to daisy chain my slow and fast movers, I don’t need two separate systems.” That’s not just less money up front, it also means just one system to maintain in the future.
There may be other benefits: If you’re using high-density storage as part of the delivery system, you may be able to build a smaller distribution center; there is also a quality factor, since a robot is unlikely to put an item in the wrong tote; finally, automation raises the profile of a facility, which may allow it to attract and retain the type of labor it does want.
Schaefer is not alone in attacking the robotic piece picking problem. Axium has also rolled out a piece picking robot for one 3PL customer and says it has other systems in pilot. The real issue for these companies is selling them to a skeptical market. Or, as Schmit put it: “Why can a company in Brazil justify the system, but not a company in North America?”
Moments later, he answers his own question. “You need systems integrators who have the size and commitment to make this kind of system work,” he says. Clearly, Schaefer fits that bill. But, he adds, you need two to tango. “Robotic solutions also need customers that have the vision to see what’s achievable.”