Robot hand with the softest touch

Engineers at Cornell University have developed a robotic hand that uses light sensors to detect the shape and texture of objects.

Soft robotics differ from traditional robots made from rigid parts, and powered by joints and motors. It is a growing field of research with products such as soft robot grippers already used in factories to sort and pack food like tomatoes or baked goods. Now engineers at Cornell University have developed the technology to create a touch-sensitive robotic hand.

The hand uses an array of optical detectors to sense shape and texture through touch, capabilities that have been tested on the hand’s ability to determine a tomato’s ripeness. The prototype soft prosthetic hand is equipped with a system of photosensors and LEDs embedded in flexible tubes. These sensor-laden tubes can detect force, curvature and elongation. And, because they are built into the hand – they are on the inside of the ‘body,’ rather than on the surface like most robots – the forces are detected as they are transmitted through the tubes. Huichan Zhao, a doctoral candidate in mechanical engineering at Cornell who is the lead author of the research team explains why this is important, “Our human hand doesn’t function using motors to drive each of the joints but is soft with a lot of sensors … on the surface and inside the hand” she says. Soft robotics provides a chance to make a soft hand that is more close to a human hand.”

In April last year we wrote about a prosthetic hand that connects to the wearer’s smartphone, enabling them to control its movements via an EMG sensor. With Zhao estimating that the soft robotic hand can be manufactured for under USD 50, it’s clear this development paves the way more sensitive prosthetics. Could this technology also be used to develop more sensitive soft robots able to work safely near humans?