Search By Topic The Green Supply Chain Distribution Digest
Supply Chain Digest Logo

Category: Supply Chain Trends and Issues

Supply Chain News: Patent Filing from Amazon Suggests Robotic Piece Picking will not be Easy


Robots Dropping Items could be Big Problem, Amazon  Says, Describing a Machine Learning System to Address

Nov. 19, 2019
SCDigest Editorial Staff

For at least a couple of years, Amazon held a robotic piece picking challenge, bringing teams in from across the world to test their robot's ability to pick an varied set of items from static shelving and place them in a tote.

Supply Chain Digest Says...

What to do? The Amazon patent filing describes a new type of control system that uses machine learning to self-adjust.

What do you say?

Click here to send us your comments
Click here to see reader feedback

Not surprisingly, it appears Amazon itself is working hard on robotic piece picking – and says it is going to be very challenging to get it right.

That after three researchers at Amazon filed a US patent for what it calls an "adaptive perception for industrial robot systems," but the filing almost exclusively deals with challenges with piece picking in distribution centers.

The filing notes that while it might be a smart choice to use robots for picking for productivity reasons or to avoid repetitive motion type injuries, "it is critical that any such automated system operates at a very high rate of success."

It notes that as a robot arm picks a product from storage for placement in a tote or on to a conveyor, the robots frequently drop items in trying to complete the task. That in turn is very disruption to the workflow of the DC, Amazon says.

The patent filing envisions a scenario with different robots working to complete different order picking tasks, as seen in the graphic below.

But getting robots to do that consistently is a challenge, in large part because of the variance in item shapes and sizes. Different items can also require different sorts of tools on the robot arm and varied approaches to the object. All of that means adequately teaching the robot how to complete all the picks is very difficult.


Amazon's Robotic Picking Scenario in New Patent Filing


When failures occur, the patent says, robot technicians can usually make adjustments to correct the problem – for that specific item. But even that can get tricky, as perhaps one robot has a problem with an item and another robot of the same model does not.

What's more, this results in a "manually intensive process and requires a significant amount of experimentation," the report says, making it impractical for DCs processing thousands of different SKU types.

(See More Below)



What to do? The Amazon patent filing describes a new type of control system that uses machine learning to self-adjust. The robots would be provisioned with a variety of sensors and tracking capabilities so that the control system knows the action involved, what item was being picked, some information about the drop and more.

Somehow with that data, the control system will learn what went wrong, and make adjustments the next time the situation is discovered.

The control system will have other smarts too. For example, it will track changes in "luminosity" (the amount of light it the DC at any given point of time) and potentially make other adjustments because its vision system is affected by the amount of light.

Like most patent filings, the bulk of its 23 pages is very technical but if you want to wade through it, it is available here: Amazon Piece Picking Robot Control System Patent Filing

What do you think Amazon's thoughts on robotic pierce picking? Let us know your thoughts at the Feedback section below.




Follow Us

Supply Chain Digest news is available via RSS
RSS facebook twitter youtube
bloglines my yahoo
news gator


Subscribe to our insightful weekly newsletter. Get immediate access to premium contents. Its's easy and free
Enter your email below to subscribe:
Join the thousands of supply chain, logistics, technology and marketing professionals who rely on Supply Chain Digest for the best in insight, news, tools, opinion, education and solution.
Home | Subscribe | Advertise | Contact Us | Sitemap | Privacy Policy
© Supply Chain Digest 2006-2023 - All rights reserved