Skip to Content
Uncategorized

Meet Domo

It just wants to help.

When Aaron Edsinger was 12, he grew impatient with the family TV, which had a clunky dial and no remote control. So he built a mechanical hand that let him change channels from the couch.

Aaron Edsinger, SM ’01, PhD ’07, works with Domo, a prototype household robot.

Now Edsinger, SM ‘01, PhD ‘07, is the lead researcher behind Domo, a robot designed to assist elderly and disabled people with tasks like shelving groceries. “The future for robots involves adapting to changes in surroundings,” says the 34-year-old postdoctoral associate at MIT’s Computer Science and Artificial Intelligence Lab.

Robots’ core elements include sensors that transmit signals to register stimuli, actuators that convert energy into motion, and software that controls actions. For factory-floor robots, those actions are repetitive, but Domo is no one-trick pony. “I can work with Domo for a half-hour and never do the exact same thing twice,” Edsinger says. He and his collaborators developed algorithms that let Domo place unfamiliar objects on a shelf or in a cup. “It’s not that robots have never adapted in real time before,” he says. What’s new is how broadly Domo adapts on the fly.

Domo inherited much from two MIT-designed forebears: Kismet, which was created to interact with humans, and Cog, which was designed to learn about objects by poking them with its hand. (Edsinger worked on both; Cog was at the heart of his master’s thesis.) Domo has Kismet’s face-tracking capabilities, so its enormous blue eyes can follow humans and encourage interaction. From Cog, Domo gained the framework for its arms, as well as the shock-­absorbing actuators, called series elastic actuators, used in 22 of its 29 joints.

Domo responds primarily to verbal commands but registers physical cues as well. Say “Domo, here,” and the robot fixes its gaze on the speaker, extending a hand to receive objects. Grab Domo’s arm while it reaches for a shelf, and it stops. A dozen computers help the robot interpret its surroundings. The data flow through an Ether­net connection to a main computer, which controls Domo’s body.

Human-robot collaboration can be of great value, Edsinger says: “A person may have great planning and perception skills, but maybe they don’t have great physical ability. The robot and person can work together as a team.”

Keep Reading

Most Popular

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.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

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.