Close

GWU Pet Project Means Your Dog Really, Really Wants a Robot

Students at George Washington University are teaching their PR2 to be man's best friend's best friend

2 min read
GWU Pet Project Means Your Dog Really, Really Wants a Robot

Teaching robots to S.C.O.O.P. P.O.O.P. is fine for humans with dogs, but the dogs themselves couldn't care less what happens to their, uh, leavings. What dogs do care about is going for walks, playing fetch, and getting fed, and in an effort to appeal to the canine user demographic, students at George Washington University are hard at work teaching their PR2 robot to be your dog's new best friend.

GWU, in Washington, D.C., offers a course called Autonomous Robotics, where students learn how to program robots to perform occupational tasks. First they get everything working in simulation, and then they move on to a real live robot. This semester, most of the students in the course chose to work on pet-related tasks, including taking a dog for a walk, playing fetch with a dog, and bringing a dog food. No cat people at GWU, I guess.

As with just about everything in robotics, these tasks sound simple, but are actually fairly complicated. Let's start with dog walking: Anyone with a dog will tell you that taking the dog for a walk usually means holding onto a leash while the dog runs around doing its own thing. But if you're trying to get a robot to do the job, you have to treat the dog as an obstacle to be avoided (as robots tend to do to keep from colliding and "crushing hu-mans"). As a result, getting a robot to manage a leashed animal ended up being more about the dog walking the robot than vice versa.

The challenge behind the dog food delivery scenario is teaching a robot to transport an object with a potentially unstable center of mass, like a big bag of dog food that is full of kibbles n' bits that shift around as the robot moves. The simplest version of this system might be a box with a heavy ball inside it: the robot has to be able to compensate for the fact that the ball rolls around. The cool thing about this project is that it's adaptable to much more than dog food; moving unstable objects is one of those things that robots are going to have to be able to do reliably in order to be useful around the house, whether it's bringing your K-9 some chow or serving breakfast in bed like in my future robot fantasies.

But, back to the pet-related tasks... My advice? Just retrofit PR2 with an automatic tennis ball launcher and call it a day.

[ GWU ] via [ Willow Garage ]

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman
LightGreen

This article is part of our special report on AI, “The Great AI Reckoning.

"I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less