Quantcast
Channel: Artificial Intelligence
Viewing all articles
Browse latest Browse all 1375

A new robot helper could make daily chores astronomically more fun

$
0
0

dirty dishes dirty apartment plates sink nasty gross smelly bad

The robots are coming, and that's not necessarily a bad thing. In fact, a robot could be the best roommate you've ever had.

New research has imbued a robot arm with a special skill: the ability to offer a helping hand doing the dishes.

The researchers used body tracking technology to teach robots how to pay attention to and react to a human's body movements when working together around the house.

This has been a big problem for scientists trying to develop robots that will collaborate with humans — they have trouble adapting to our personal style. "We want robots to follow our lead, or at least plan their actions with an awareness of ours," Bilge Mutlu, of University of Wisconsin-Madison told MIT Technology Review.

Mutlu and collaborators from the University of Madison-Wisconsin and the University of Washington published the study in Robotics Science and Systems Online Proceedings.

If you're unloading a dishwasher with another person, you will speed up or slow down how quickly you hand off each dish according to your partner's readiness. Robots helping with such a task should learn that same awareness, so as to avoid smacking you with a plate while your back is turned.

To teach the robot, researchers recorded a pair of volunteers unloading dishes in different ways. A Microsoft Kinect documented the movements of volunteer's body joints. The program then "watched" this interaction, taking note of the thousands of joint readings from the Kinect and training them into an algorithm.

robot handoff 1The team then analyzed the movements of eight teams and found that the "giver" handing the dishes off to the "receiver" monitored and adapted to the "receiver's" pace by either pausing or slowing down their speed until their partner was ready.

They trained a Kinova Mico robot arm with their program and put it to the test in the same scenario, shown below. The robot's algorithm was able to predict when the "receiver" was ready for a dish about 90% of the time, according to MIT Technology Review, by tracking the partner's movements in real time and adjusting its speed accordingly.

robot handoff 2While the idea of an intuitive, learning robot sounds extremely useful, the researchers note that this success only applies to this one task of unloading dishes. In the future, as reported by MIT Technology Review, they hope to apply it to other more useful chores such as unloading groceries, handing off tools, or guiding someone through a physical therapy session.

Join the conversation about this story »

NOW WATCH: University of Michigan just opened their $10 million city for driverless cars


Viewing all articles
Browse latest Browse all 1375

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>