Skip to Content

Forget Drones—Google Explores Robot Delivery

Legged robots from Boston Dynamics can navigate a home, and even deliver a parcel, using advances in manipulation and vision.
December 8, 2016

The nimble-legged robots under development at a secretive Google subsidiary are getting ever more capable and clever.

At a conference in Barcelona this week, Marc Raibert, the CEO of Boston Dynamics, which specializes in dynamically balancing legged machines, demonstrated some of the progress his researchers have been making.

Raibert demonstrated Spot Mini, the company’s latest four-legged robot, which is about the size of a large dog. Boston Dynamics has previously shown videos of Spot Mini operating in a mocked-up home—climbing stairs, opening doors, and even emptying a dishwasher using its gripper. The robot features a neck-like appendage and gripper that enables it to do simple, but potentially useful, manipulation tasks.

The robot is partially automated. A Boston Dynamics engineer steered a Spot Mini onto the stage during Raibert’s talk at the Neural Information Processing Systems conference. But the robot figured out for itself how to perceive and navigate the steps up to the stage, and then, once given the command, located and picked up a can from a table.

Legged robots could potentially be better than wheeled bots at navigating messy human environments, although the research robots under development at Boston Dynamics remain prohibitively expensive for now, some costing more than $1 million.

Boston Dynamics has built a reputation for developing robots capable of walking and running, even across treacherous ground using dynamic balance; that is, by constantly moving to maintain stability. The company has honed the technique over many decades to produce several stunning machines  (see “The Robots Running This Way”). It makes a much larger quadruped, called Big Dog, which has been tested as a military pack mule, as well as a humanoid, Atlas, which took part in a robot rescue contest organized recently by the Defense Advanced Projects Research Agency (see “Why Robots and Humans Struggled with DARPA’s Challenge”).

As Boston Dynamics explores potential applications, it’s clear that manipulating objects while balancing this way will be a key focus. “Mobile manipulation is our next grand challenge,” Raibert said during his talk.

Raibert showed several videos of Spot Mini doing tasks involving manipulation, like opening weighted doors with oddly shaped handles. He also showed experiments with having the robots carry packages up to the front door of different homes around Boston. “Many people are talking about drone delivery,” Raibert said. “So why not just plain legged robots?”

Boston Dynamics does not currently use machine learning—a key focus of the NIPS conference—in its robots, but Raibert said this could change, as the technique could offer a way to automate some of the programming required for dynamic balancing as well as other tasks.

Raibert also offered a few clues as to the company’s future plans, saying that the company is continuing to make its robots lighter and more energy-efficient. This will likely include new, leaner versions of Spot and Atlas.

Boston Dynamics was one of several innovative robotics companies acquired by Google in 2013, as the company sought to build out a cutting-edge robotics effort. This overall project has floundered, however, with key team members leaving amid reports of a lack of overall direction. Early this year, it was reported that Google hoped to sell Boston Dynamics as it worked to focus its robotics research on areas that might have more immediate commercial applications (see “Google Hasn’t Given Up on Robots”).

Keep Reading

Most Popular

Scientists are finding signals of long covid in blood. They could lead to new treatments.

Faults in a certain part of the immune system might be at the root of some long covid cases, new research suggests.

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.

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.