A New Method Might Finally Offer Tactile Sensation in Robotic Fingers

And it achieved nearly 100% accuracy.
Brad Bergan
A human hand touching a robotic one.Blue Planet Studio / iStock

Human senses are hard to replicate.

Despite this challenge, roboticists from around the world have spared no effort in the pursuit of novel robotic systems that can artificially emulate the familiar sense of human touch, and researchers just developed a new tactile sensing method that could enhance robotic materials composed of soft materials, according to a recent study shared on a pre-print server.

And it's a landmark feat of engineering on the road to a human-like sense of touch in robotic limbs.

Robotic finger with human-like touch uses proprioception

In the last several years, roboticists have confronted the problem of artificially replicating the human sense of touch via fabricating increasingly advanced and life-like bionic limbs, in addition to humanoid robots, employing soft materials, instead of the conventional, hard, and metallic structures of popular robots. But while soft material-based robotic hands have commendable advantages in skin-like texture, this design also lacks the ability to collect a wide spectrum of sensory input. And replicating complex biological mechanisms that make human touch and its capacity to offer information on objects without other senses has become a substantial challenge.

Cheng Robot Finger
The finger device, the strain sensor, and the actuator, from left to right. Source: Cheng, et al. / Beihang University

This is why researchers at Beihang University in Beijing developed a new sensing technique that involves soft material-based robotic fingers. Called proprioception, this biological mechanism is what enables mammals to perceive a degree of situational awareness, including position, and movement. "The idea behind our recent paper is based on the proprioception framework found in humans, which is what determines our body position and load on our tendons/joints," said one of the researchers, named Chang Chen, in a TechXplore report. "Think about when you put a blindfold on and cover your ears, you can still feel your hand posture, arm position, or how heavy a grocery bag is; this ability is known as proprioception."

Most Popular

"We have been working on a prosthetic hand research project and we are looking for ways to address the lack of sensory feedback in existing prosthetic hands," added Chang Cheng, in the report. Earlier attempts saw robotics researchers working without correlating proprioception with the sense of touch. This isn't totally surprising, since proprioception doesn't directly contribute to precise responses, which is likely why humans don't use it to identify the texture of surfaces on objects. But while industrial-grade sensors are far more sensitive than human proprioceptors, implementing them in robotic fingers might open the door for researchers to learn more than ever before about highly precise tactile feedback.

Cheng Robot Finger Flexes
The new finger method flexing with different tendon strains. Source: Cheng, et al. / Beihang University 

The touch-sensitive robotic finger could see a major nanotech upgrade

Cheng and his colleagues developed a prototype system that involves a linear actuator, a strain sensor, a tendon (cable), and a soft robotic finger from an earlier project of theirs. "The tendon connects the finger to the actuator and the strain sensor is installed in the middle of the tendon," explained Cheng, in the report. "When the actuator is driven, it pulls the tendon, which causes the finger to bend/straighten, and the strain on the tendon changes accordingly. When the finger touches different objects, the sensor would output series of strain signals that characterize the touched objects." In short, the new method starts with a reading from the sensor, then employs machine learning tools to decode the rigidity and texture of an object or surface while the robotic finger is in contact with it.

And this technique could read stiffness and texture with unconscionably high levels of accuracy, at 99.7% and 100%, respectively. This is a landmark accomplishment in the pursuit of human-like tactile sensation via robotic hands, even if it's only with a single finger. As Cheng's team further develops the new technique, they will collaborate with a well-known nanotechnology lab to develop less pricey tactile sensors that enable the sensing of force and torque. It's not a fully functional bionic limb yet, but it's one of the engineering feats that could bring them into reality.

message circleSHOW COMMENT (1)chevron