Fishing keys from the bottom of a bag or picking up soft fruit are pretty basic tasks for us humans. But they’re not so easy for robotic hands. Recognising and grasping different objects usually requires complex programming and processing power.
at Nanyang Technological University in Singapore and his colleagues have developed flexible skin-inspired touch sensors that store tactile information. The sensors work like our haptic memory, which can store impressions of touch sensations in the brain after the stimulus has stopped.
Chen says the sensors could store information to help robots recognise their environment and moderate their grip strength to, and to be delicate to avoid damaging things like fruit. This frees the robot’s main processors for other tasks. What’s more, he reckons they could provide robots with personalised secure activation, such as a handshake from a particular person.
The sensors comprise a pressure-sensitive layer that detects changes in electrical resistance when force is applied. Beneath this layer is a thin-film memory device that records these changes to form a digital impression of any pressure. The researchers showed the sensors can retain such information for about a week, though data can also be erased by applying a voltage.
Not everyone is convinced that giving memory to artificial skin is useful. “The need for such memory in robotic and prosthetic applications is not clear,” says, a roboticist at the University of Bristol, UK. “Typically, one wants to have information about a touch event as quickly as possible. Having an embedded record of touch isn’t so immediately useful.”
at the Johannes Kepler University Linz in Austria thinks this approach has potential. “It may be interesting when gripping complex-shaped objects to know the forces exerted on them,” he says. “Imagine gripping soft objects, such as strawberries: here it is essential to know the contact forces of a robotic gripper.”
Without this tactile memory, our future robot chefs are going to squash a lot of fruit.
Journal reference: Advanced Materials, DOI:
Image credit: Benny J. Johnson/Getty
More on these topics:
This entry passed through the Full-Text RSS service – if this is your content and you’re reading it on someone else’s site, please read the FAQ at fivefilters.org/content-only/faq.php#publishers.