UT Austin researchers develop robotic hand gentle enough to pick up fragile items

UT researchers develop robotic hand

A robotic hand developed at UT Austin can pick up the most fragile items like potato chips or eggs without crushing them. FOX 7 Austin's Jenna King spoke with the team behind the project about how it works.

AUSTIN, Texas - A robotic hand developed at UT Austin can pick up the most fragile items, like potato chips or eggs, without crushing them.

Researchers say this breakthrough could have a big impact across multiple industries, from health care to manufacturing.

What they're saying:

"We want robots to pick things up, right? But robots don't currently have a sense of touch," said Lillian Chin, senior author on the project.

Until now.

From a potato chip, to an egg, to a raspberry, items that would normally crumble or crack can now be picked up safely thanks to a new robotic hand developed at UT Austin. The technology called Fragile Object Grasping With Tactile Sensing allows the robot to "feel" what it’s holding.

UT Austin robotic hand demonstration

UT Austin researchers have developed a robotic hand gentle enough to grasp fragile items like eggs and even a potato chip.

"The new thing, especially in this, is the force detection, like how much force you're putting onto an object, so you don't squish it. And then the slip detection. So right when you're holding something, like being able to tell that it's slipping out of your grasp. So those are the new features," said Chin.

The flexible fingers are modeled after fish fins and created using advanced 3d printing. As the fingers prepare to grab an object, the air channels inside of them shift, which causes a change in air pressure.

The pressure changes detected provide real-time force feedback to the robot to let it know whether the object is slipping.

"What my lab has done is we have this technique to put touch sensors in basically anything. So, we 3D print something, and then we put these empty air channels inside of the structure. So that way, if you like move the structure around or compress it, those air channels will feel a force and then you can measure that," said Chin. 

By the numbers:

Researchers tested the grippers on 31 different objects, including potato chips, raspberries, jam jars, soup cans and apples.

The system successfully picked them up about 92 percent of the time outperforming traditional robotic grippers.

It also detected 93% of slips with perfect precision, meaning it never falsely identified a slip.

"I don't know too many other robots who like have this like slip sensing capability. And then ours are really high durability. So, we've been using this for over 200 hours of experiments, whereas like these state-of-the-art camera ones. They can only work for an hour before they start having issues," said Chin.

What's next:

Researchers say the technology could one day be used in food processing to reduce waste, in health care to handle medical tools or fragile biological samples, and in manufacturing for delicate items like electronics or glass.

"I see it as being anywhere that we care about fragile and delicate objects, which is pretty much anywhere there's humans involved. This technology could be used," said Chin.

The team has also released their designs and algorithms to other engineers to build on the technology.