Electronic Skin Gives Robots a Sense of Touch

Zothecula writes "Providing robots with sensory inputs is one of the keys to the development of more capable and useful machines. Sight and hearing are the most common senses bestowed upon our mechanical friends, but even taste and smell have gotten a look. There have also been a number of efforts to give robots the sense of touch so they can better navigate and interact with their environments. The latest attempt to create a touchy feely robot comes from the Technical University Munich (TUM) where researchers have produced small hexagonal plates, which when joined together, form a sensitive skin."

Comments

Like I'd take that chance.

[sgt scrub]

Nothing would be worse than a love bot that can say you make her skin crawl.

Re:Like I'd take that chance.

[Yvan256]

That's not a love bot, that's a snake bot!

Re:Like I'd take that chance.

[fuzzyfuzzyfungus]

An experience with a lovebot that possesses actuators of sufficient power to be interesting, without a touch-sense feedback system might give you considerable insight in to things that are worse.

This post brought to you by the phrase "TRAUMATIC DEGLOVING LESION OF PENILE AND SCROTAL SKIN"

Star Trek meets reality

[ArhcAngel]

We don't have competent AI yet but when we do they'll be able to feel the people they grope.

Re:Star Trek meets reality

[SniperJoe]

Hrm... incompetent, robotic and groping. That sounds an awful lot like the TSA!

Re:Star Trek meets reality

[Jason Levine]

Maybe this research was funded by the TSA to replace their already low paid employees. Coming soon to an airport near you! "Strip naked and get on the probulator!"

Re:Star Trek meets reality

[thePowerOfGrayskull]

Hrm... incompetent, robotic and groping. That sounds an awful lot like the TSA!

Not to mention most teenagers.

interesting approach

[rbrausse]

I like the idea of using infrared for simulating skin sensivity. At first I thought hard modules are a bad way for creating a artificial sensual skin but using a solution without the need of direct contact could be useful (e.g. in highly damageable environments).

This will be a tricky one...

[fuzzyfuzzyfungus]

Biological systems have an impressive advantage in that their complexity is cheap(in fact, given that they have to grow it, and thus vascularize or otherwise provide for nutrient and gas exchange, it might even be that structural complexity is their only option in many cases)...

Even a small, boring, bug might have several thousand sensory hairs, each capable of detecting mechanical disturbance(possibly even the direction of that disturbance). With machines, by contrast, semiconductor fabrication technology has brought the cost of small areas of extremely high complexity down(CCDs, MEMS components, microprocessors); but the cost and difficulty of having thousands to hundreds of thousands of distributed sensors(and a wiring harness to connect them all...) is still quite considerable.

Re:This will be a tricky one...

[newcastlejon]

I'm reminded of an argument/discussion I once had with a friend of mine who read environmental *ahem* engineering about the differences between biological and synthetic systems, specifically plants vs. solar panels. He argued that trees are inherently superior to solar panels because they grow unaided from such a small thing as a seed. I didn't dispute this point, but couldn't help point out that you can't make a tree that's identical to the one next to it, even down to the metre or centimetre scale, much less micrometre.

One can't deny the merits of a living, self-repairing and self-assembling system like a tree or an insect, but if you want to have more than one copy you have to build it yourself. Robots with tactile senses are naturally going to have a big advantage of those that don't, but I think the best way to go about it would be to follow the same route as SKYNET supposedly did: build a metal robot, but give it a flesh coating. We can build large moving machines more efficiently than nature and given the manufacturing technology we might be able to do the same for smaller-scale parts too, but for the moment nature really does beat technology when it comes to producing things that function at small scales in bulk. Take image sensors as an example: we could probably make an artificial retina that's much more sensitive than the real thing but to produce a lot of them it's probably better just to settle for the squishy kind and replace the problems with silicon processing with petri dishes and simple nutrient broth.

good

[marcusi83]

It would be incredible

Might be nice for safety but the rigidity makes it

[divisionbyzero]

useless for developing an android. Still interesting but comparing it to skin is totally misleading. It's more of a compact sensor of the sort found on robots than something "organic".

I'd use a flexible capacitive sensor.

[Ptolom]

I you asked me to design a robot's skin I'd probably try to make some sort of flexible touchpad. The bending would seriously distort readings though.

Re:I'd use a flexible capacitive sensor.

[fuzzyfuzzyfungus]

Don't think of the distortions caused by bending as 'distortions' think of them as "signals caused by exactly the occurrences I am trying to measure"...

A flexible surface that can measure its own deformation state with reasonable speed and accuracy is a touch sensor, albeit not the most sensitive...

Terminators need skin to use time travel!

[Joe_Dragon]

as they need to skinned with living mater.

Killbots

[ThatsNotPudding]

with a human touch.

NASA? 6-7 years ago?

[Eg0r]

... they even put up a video of a Ballerina dancing around a giant penis... I think at the time it was voted one of the "worse jobs in science" or something...