Robotic Shape-Shifting Gripper Picks Up Anything

A seemingly simple task for humans—picking up objects of various shapes —can be quite complex for robots. A new shape-shifting technology could soon change that.

A new robotic gripper consists of a bag filled with coffee grounds or other grain material. Initially, the bag is soft and malleable and can conform to any object. Once the gripper is in place, an attached vacuum pump sucks all the air from the bag, solidifying its grip and allowing a user to pickup the object.

When air is allowed to flow back into the bag, the gripper softens and the object is released.

“The ground coffee grains are like lots of small gears,” study co-author Hod Lipson, a mechanical engineer at the University of Chicago, said in a statement.“When they are not pressed together, they can roll over each other and flow. When they are pressed together just a little bit, the teeth interlock, and they become solid.”

While seeking the right material for their gripper, the researchers experimented with everything from rice to couscous and even ground-up tires. Coffee beans were found to provide the right combination of light weight and good interlocking ability.

So far the researchers have made a golf ball-size gripper filled with coffee grounds and shown its ability to pick up a variety of small objects, including M&Ms, bottle caps and foam ear plugs.

The size of the device could be scaled up or down, depending on its use. “You could potentially lift up a car with something like this if you made a big enough gripper,” said Eric Brown, a researcher also at the University of Chicago and one of the study authors.

Potential applications for this universal gripper include prosthetic arms, search-and-rescue robots, and robots that sweep for mines or improvised explosive devices, Brown told TechNewsDaily.

Typically, robots sport claw devices meant to mimic humanhands. “The hard part about that is you have a lot of hinges and joints in the human hand,” Brown said.

“As humans, we’ve learned how to control all of those joints very well,” Brown said. “From a robotics standpoint, that’s a lot of objects to control, so you need a lot of sensors and a computer.”

This simpler approach came about through research funded by the Defense Advanced Research Projects Agency. Researchers from the University of Chicago, Cornell University, and iRobot Corp.were challenged to develop robots out of soft materials.

Brown and his colleagues detail their development online in today's issue (Oct. 25) of the Proceedings of the National Academy of Sciences.

Michelle Bryner
Michelle writes about technology and chemistry for Live Science. She has a Bachelor of Science in Chemistry from the Salisbury University, a Bachelor of Chemical Engineering from the University of Delaware and a degree in Science Journalism from New York University. She is an active Muay Thai kickboxer at Five Points Academy and loves exploring NYC with friends.