From Goods to Picker to Piece Picking Robots, Software is the Key

Most logistics professionals know robots are starting to have a huge impact on distribution operations.

They may not know artificial intelligence and machine learning is making the productivity gains possible.

Supply Chain Digest Says... AI has been married with vision systems and new advances in "grabber" technology to make great advances in robotic piece picking. What do you say? Click here to send us your comments Click here to see reader feedback

According to an article in the Wall Street Journal, XPO Logistics is deploying 5,000 AI-equipped robots to deliver bring shelves full of products to workers. Rakuten Super Logistics, a division of Japanese on-line retailer Rakuten, is deploying robots to retrieve bins filled with apparel, electronics and other products from shelves and deliver them to workers who pick individual items. Gap stores are using AI-aided mechanical arms to help sort clothing orders.

The key to these new age robots is the software, which analyzes data in real time and leverages machine learning over time to decide the best order for picking a series of items or where to store goods based on demand.

The WSJ article also says machine learning is being used to teach autonomous mobile robots how to map out DC environments and adapt to changes in layout or stocking locations.

Another play here are so-called collaborative robots, which unlike traditional industrial robots are designed to work safely right alongside human beings. While the "killer app" for this style of robots in distribution has yet to emerge, there appears to be a market for using such robots to meet seasonal demand spikes.

One form of these robots does seem to be gaining some traction: smart carts that move autonomously to the next picking location, leading the human being to the right location, then using a light system to identify the SKU and quantity to pick.

"The big efficiencies come from machine learning," Brad Jacobs, XPO's chief executive, said in an interview last year when the 3PL announced it was deploying AI-equipped robots from Singapore-based robotics firm GreyOrange Ltd. across its North American operations.

XPO says that so far, it has has seen picking productivity increase four times from robot deployment.

Many of these robots are similar to those developed by Kiva Systems, which was acquired by Amazon in 2012. Those orange Kiva robots largely ushered in the "goods to picker" concept in distribution automation, an approach that has spread to other technologies used in DCs such as shuttle systems.

(See More Below)

The aforementioned Rakuten began using 40 autonomous mobile robots from a company called inVia Robotics at its Las Vegas facility before this year's peak season.

"We're probably picking 50% to 70% more orders in the same time frame than if we just did the human pick," Mary Marriott, the company's vice president of operations, told the Journal.

The company is also using these robots in a goods to picker fashion.

"In beta testing we told the workers the robots are there to support you and help you do your job more efficiently," Marriott also says. "The team members took it in stride and really embraced having the robots in that facility, especially when they saw it made their job easier."

AI has been married with vision systems and new advances in "grabber" technology to make great advances in robotic piece picking, sort of the Holy Grail of DC automation.

In recent years, Amazon has been holding a global "robotic picking challenge," where teams from across the world compete on the best design for picking an add assortment of items from bins.

The money keeps rolling in to many of these start-ups. As we reported last week, RightHand Robotics has just raised $23 million more in investment to support its efforts to develop piece picking robots for distribution centers.

That investment came from private equity firm Menlo Ventures, Google parent ABC company, and other firms. That brings the total raised by the four-year-old company to $34.2 million.

SCDigest has covered RightHand Robotics for several years now – the Boston area company incorporates artificial intelligence and cameras that perceive depth and color with a robotic arm that uses a unique polymer gripper with a suction cup to pick up objects, say out of a static shelving location. Beyond the advances in the physical dexterity of the new age robots, it is the use of AI that is really changing the game.

How fast do you believe robots will take over DC actvities? What are the key issues? Let us know your thoughts at the Feedback section below.