Shyam's Slide Share Presentations

VIRTUAL LIBRARY "KNOWLEDGE - KORRIDOR"

This article/post is from a third party website. The views expressed are that of the author. We at Capacity Building & Development may not necessarily subscribe to it completely. The relevance & applicability of the content is limited to certain geographic zones.It is not universal.

TO VIEW MORE CONTENT ON THIS SUBJECT AND OTHER TOPICS, Please visit KNOWLEDGE-KORRIDOR our Virtual Library

Tuesday, May 21, 2013

Ikea Bot Assembles Furniture Using “Common Sense” 05-22


IkeaBot Assembles Furniture Using “Common Sense”

by NANCY DUVERGNE SMITH
Assembling popular Swedish furniture may help the masses, but it’s only a hint of what goes on in MIT’s Distributed Robotics Laboratory. How the bots do it is the breakthrough. According to a recent IEEE Spectrum article, the bots are fully autonomous and need no human help to whip together a Lack table in 10 minutes. The magic is in the software and the grippers—and that magic can be applied to industrial-scale problems in manufacturing.
Ross A. Knepper, a postdoctoral associate, is leading the effort to teach a team of commercially available KUKA youBots to assemble the furniture. In an earlier life, he created motion planners that drive Mars rovers, unmanned military vehicles, and a personal home-assistant robot called HERB.
With the Ikeabot, Knepper is tackling a key problem in robotics with savvy algorithms.
“A lot of problems in factory automation are similar to the problem in Ikea furniture assembly,” says Knepper. “There are many robots in factories but they perform very simple functions. In the future, we want robots that can move around in the factory and interact with people…so they can be treated as teammates, not just tools.”
Knepper is writing code that creates the kind of common sense that allows humans to work side by side intuitively. “If you imagine two people assembling furniture together, they can infer what the other is doing—they don’t have to explain it. [The IKeabots] are trying to infer how parts fit together and the logical order of assembly.”
Using a natural language feature, the robots can ask for help. If they can’t reach a part, for example, they find a human and ask that the part be handed to them, and then they continue to work.
Space requirements have guided much of robot research in the past few decades, Knepper says. In space, robots need a higher order of intelligence to solve problems and work independently. The payoff may be closer to home though—on the factory floor. Using intelligent robots could help rebuild manufacturing and create jobs in the US. “We will need highly skilled people to operate the robots and robots and humans can trade off jobs,” he says. “You can have a much more efficient process.”
What’s next for the Ikeabot? The team is working on an Allen Wrench glove that the robot can put on and off as needed, and the future is about groups of robots working collaboratively with one another—and with people. And all that fits neatly into the Distributed Robotics Laboratory, which is headed by Daniela Rus, director of the Computer Science and Artificial Intelligence Laboratory (CSAIL). DRL is known for research in programmable matter and distributed robotics. In fact, the lab’s robots have many talents: they can end a garden, bake cookies from scratch, fly in swarms to perform surveillance functions, and dance with humans.
Want the details? Download “The IkeaBot: An Autonomous Multi-Robot Coordinated Furniture Assembly System,” which was nominated for Best Automation Paper at the International Conference on Robotics and Automation (ICRA)in Karlsruhe, Germany, May 2013.