Researchers at the National University of Singapore (NUS) are working hard to teach robots how to feel. And, no, not in the sense that their subjects are opening up emotionally about the state of robotics, circa 2020. Instead, they’re learning to feel — as in, to use sensory identification to detect touches and identify the shape, texture, and hardness of objects, much as we humans do.
What the researchers have developed is an artificial skin for robots they claim is able to detect these touches more than 10,000 times faster than the (already impressively rapid) human sensory nervous system. While it’s still early days for the research, it could nonetheless open up a plethora of new applications for future robots.
“Enabling a sense of touch in robotics could lead to many new use cases, and in general, allows for safer human-robotic interaction by improving the robot’s perception of its environment,” Mike Davies, director of Intel’s neuromorphic computing lab, which developed the Loihi neuromorphic research chip used to power the robotic skin, told Digital Trends. “For example, robot-assisted surgery technology would greatly benefit from more sensitive and faster tactile sensing. In the factory, challenging manufacturing tasks such as soft material assembly and spring detangling will be impossible to automate until industrial robots can achieve sufficient fine motor dexterity with tactile sensing.”
In an initial experiment, the researchers used a robotic hand wearing their artificial skin to read Braille. It was able to achieve this with more than 92% accuracy, while using significantly less power than alternative approaches. The team built on this work by looking at how a combination of vision and touch data could be used to classify objects. The results were revealed at the ongoing Robotics: Science and Systems conference, which runs through this week.
“NUS plans to further develop this robotic system for applications in the logistics and food manufacturing industries, where there is a high demand for robotic automation — especially moving forward in the post-COVID era,” Davies continued. “These benefits can be compounded by adding additional sensing capabilities, such as hearing and smelling.”
It’s not quite at the level of Terminator-style cyborgs, yet. But advances such as this certainly have the potential to help robots advance to the next step in their evolution.
- Intel Arc Pro is real — three new workstation GPUs revealed
- Intel has 500 bugs to fix in its next supercomputer chips
- AMD accidentally revealed a killer new feature to rival Nvidia’s RTX Voice
- These weird Intel Raptor Lake benchmarks make us question what’s coming
- Intel Meteor Lake is about to level up to a crucial stage