Scientists have developed a new type of electronic “skin” that allows robots to “feel” a variety of tactile sensations, including pork, production, temperature changes, and even stabbing sensations.
The skin is made of electrically conductive gelatin-based materials that can be molded into different shapes. Equipped with a special type of electrode, the material can detect signals from hundreds of thousands of coupling paths corresponding to different touch and pressure sensations.
Scientists said the material can be used in humanoid robots or human prosthetics, where touch-like feeling is essential, in addition to having a wide range of applications in the automotive sector and disaster relief. On June 11, they published their findings in the journal Science Robotics.
You might like it
Tactile sensing is emerging as the next major milestone in robotics as scientists try to build machines that can respond to the world in ways similar to human sensibility.
Electronic skins usually work by converting physical information such as pressure and temperature into electronic signals. In most cases, different types of sensations require different types of sensors. For example, one detects pressure and the other detects temperature.
However, signals from these different sensors can interfere with each other, and the materials they are embedded in – rubbery materials traditionally called soft silicones or elastomers – are easily damaged, scientists said.
Related: Self-healing “living skin” can make robots more human – and it looks just as creepy as you would expect
This new electronic skin uses a single type of “multimodal” sensor that can detect various types of stimuli such as touch, temperature, and damage.
While it remains difficult to reliably separate and identify the causes of each signal, multimodal sensing materials are easier to manufacture and more robust, scientists said. It is also inexpensive to produce, making it cost-effective and suitable for widespread use.
That’s convenient
To test their synthetic meat, researchers dissolved a soft, stretchy, conductive gelatin-based hydrogel and threw it into the shape of a human hand. We then equipped with various electrode configurations in our hands, and we could see what captured the most useful data from the physical interactions, and exposed them to a series of tests to find them.
This rather brutal process involved blowing up with a heat gun, thrust it in with a finger and a robotic arm, and opening it with a female.
In total, the researchers said they collected more than 1.7 million pieces of information from more than 860,000 conductive pathways in the skin. They used the data collected from these tests to train machine learning models that would allow them to recognize different types of touches when integrated into a robotic system.
“We don’t have robotic skin as good as human skin, but at this point I think it’s better than anything else,” said Thomas George Trutheran, lecturer at Robotics and Artificial Intelligence (AI) at the University of London (UCL), in a statement.
“Our methods are more flexible and easier to build than traditional sensors, allowing them to be adjusted using a variety of tasks using human touches.”
Source link
