Skip to Content
Artificial intelligence

Self-Taught Robot Is Ready to Seize Another Warehouse Job

A new robot uses computer vision and machine learning to break down pallets of boxes faster than human workers.
April 6, 2016

A keen-eyed new robot looks poised to snag an important everyday warehouse job.

Kinema Systems, a startup based in Menlo Park, California, has developed a robot capable of breaking down pallets of boxes no matter what size or shape they are or how they are packed together.

This is a routine job at thousands of large stores, warehouses, and shipping companies—the aftermath of goods making their way through the supply line to your front door.

The new robot uses a simple suction system to grab boxes, but it needs state-of-the-art computer vision and machine learning to figure out how to grab them. The machine does not need to be programmed at all—instead, it automatically calibrates itself and teaches itself how to break boxes down.

It might be trivial for a human to grab a box, but it is hard for a computer to know where one box ends and another begins, especially if the boxes are covered with labels and packing tape. And if the robot does not place its suction device at the center of a box, there is a risk it will drop it.

Sachin Chitta, founder and CEO of Kinema, says the robot uses a combination of 2-D and 3-D vision as well as learning in order to grab boxes from a pallet. The learning process also helps the robot deal with different environmental conditions such as changes in lighting. “As it picks things up, it builds a model of each box,” Chitta says. “And in the future, it can use that model to speed things up.”

The most difficult part of the process is figuring out how to grab the first box on a pallet, because everything is packed tightly together. Chitta refused to explain how the robot does that, saying it was its “secret sauce.” A new robot takes a few seconds to figure out how to pick up an unfamiliar box, and then it can grab the same box in less than a second in the future. It takes human workers about six seconds, on average, to grab a box from a pallet.

Beyond breaking down pallets, Chitta says, the company plans to use similar technology to let robots manipulate unfamiliar objects, which could be valuable for many warehouse and factory tasks.

Machines capable of unpacking pallets already exist, but they typically rely on there being a standard shape and arrangement of boxes. Efforts to break down more complex pallets using robots have mostly relied on specialized labels to identify different shapes of items.

The company’s approach “could be very valuable in contemporary logistics, where the aim is to compress delivery time from days to hours,” says Ken Goldberg, a professor at the University of California, Berkeley, who specializes in robotics and machine learning.

Goldberg adds that the information learned by a robot could be shared with other systems, even ones in different facilities, something known as “cloud robotics” (see “Robots That Teach Each Other”). “If perception and motion data can be anonymized and shared across Kinema systems, this form of cloud robotics could yield extremely robust and efficient systems,” Goldberg says.

Deep Dive

Artificial intelligence

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.