A new smart finger uses sensors to detect common materials such as glass, silicon and wood with an accuracy of more than 90 percent. This can be useful for robots in the manufacturing industry.
A newly designed artificial finger can identify different materials with an accuracy of more than 90 percent by feeling their surface. The technology can be useful for automating production tasks such as sorting and quality control.
Touch sensors that can measure, for example, the temperature of a surface, or the force acting on it, are not new. Sensors that can recognize what kind of material something is and what the roughness of a surface is are rarer.
inexplicable cat hole
Zhou Li, professor of nanotechnology at the Institute of Nanoenergy and Nanosystems in Beijing, China, and his colleagues have developed a finger that can recognize what a material is made of. The finger uses so-called triboelectric sensors for this. These sensors measure the extent to which a material can donate or absorb electrons. Furthermore, the artificial finger can determine the roughness without causing damage to the material.
The researchers tested the finger on hundreds of samples of twelve different materials, including wood, glass, plastic and silicon. They combined the measurements with artificial intelligence-based computer analysis. He learned which reaction of the sensors belongs to which material. This is how the finger took a average 96.8 percent accuracy in recognizing a material and a minimal accuracy of 90 percent for all materials.
The device consists of four small square sensors, each made of a different plastic polymer, chosen for their different electrical properties. When the sensors get close enough to an object, electrons from each square react with the surface in a slightly different way. The sensors register how the electrons react.
The researchers put these sensors in a finger-like housing. They then attached it to a computer processor and a screen that displays the name of the detected material. In an industrial environment, you could connect the processor directly to a production controller.
‘Smart fingers can help robots check whether products meet production standards in terms of composition and surface structure’, says Li. ‘Our system could also play an important role in sorting industrial material.’
If the sensor proves to be robust after many thousands of tests, then the artificial finger could indeed be well suited for tasks such as quality control in the manufacturing industry, agrees robotics researcher Ben Ward-Cherrier of the University of Bristol in the United Kingdom. But the technique would probably be more effective if you combine it with other sensors that can detect edges or friction, for example, he says.
The researchers also suggest the device could be used for artificial prosthetics. But it’s unlikely to be helpful for that, says neuroscientist Tamar Makin, a professor at the University of Cambridge in the UK. “For human-controlled technology, we don’t need this level of sophistication,” she says. “Imagine you have a prosthetic hand and you reach for a cup of coffee. You have so much life experience – and you can touch it with your intact hand – that that is enough to make a very good guess of the material you are about to touch.’