Electronic skin

jiande1

Electronic skin [Copy link]

Electronic skin that can learn from feeling "pain" could help create a new generation of intelligent robots with human-like sensitivity.
A team of engineers from the University of Glasgow in Scotland has developed an artificial skin with a new processing system based on "synaptic transistors" that mimics the brain's neural pathways for learning. A robotic hand using the smart skin showed a remarkable ability to learn to respond to external stimuli.
In a new paper published today in the journal Science Robotics, the researchers describe how they built their prototype computational electronic skin (e-skin) and how it could improve current touch-sensitive robotics technology.
Source: Bendable Electronics and Sensing Technologies (BEST) Group, University of Glasgow.
For decades, scientists have been working to create artificial skin with a sense of touch. One widely explored approach is to distribute an array of contact or pressure sensors across the surface of the electronic skin, allowing it to detect when it comes into contact with an object.

Start processing at the touchpoint
The data from the sensors is then sent to a computer for processing and interpretation. Sensors typically generate large amounts of data that can take time to be properly processed and responded to, introducing latency that reduces the skin's potential effectiveness in real-world tasks. The Glasgow team's new electronic skin draws inspiration from how the human peripheral nervous system interprets signals from the skin to eliminate latency and power consumption.

Once the human skin receives input, the peripheral nervous system begins processing it at the point of contact, simplifying it to only the important information before sending it to the brain. This reduction in sensory data allows for efficient use of the communication channels needed to send the data to the brain, which then responds almost instantly, allowing the body to react appropriately.

To build an electronic skin capable of efficient computation and synaptic-like responses, the researchers printed a grid of 168 synaptic transistors made of zinc oxide nanowires directly onto the surface of a flexible plastic surface. They then connected the synaptic transistors to skin sensors present on the palm of a fully articulated humanoid robot. When the sensor is touched, it registers a change in its resistance—small changes correspond to light touches, while heavier touches produce larger resistance changes. This input is designed to mimic the way sensory neurons work in the human body.

Processing simulates the human nervous system
In early electronic skins, input data would be sent to a computer for processing. Instead, circuits built into the skin act as artificial synapses, reducing the input to simple voltage spikes whose frequency varies depending on the level of pressure applied to the skin, speeding up the reaction process.

The team used the different outputs of that voltage spike to teach the skin the appropriate response to simulated pain, which would trigger the robotic hand to react. By setting the threshold of the input voltage to elicit a response, the team could make the robotic hand back off from a sharp jab in the center of its palm. In other words, it learned to disengage from the source of simulated discomfort through onboard information processing that mimics how the human nervous system works.

The development of the electronic skin is the latest breakthrough in flexible, stretchable printed surfaces by the Bendable Electronics and Sensing Technologies (BEST) group at the University of Glasgow, led by Professor Ravinder Dahiya.

马孔多的雨

In recent years, flexible electronic skin has been a hot topic in academia and industry, among which bionic tactile sensors used to mimic the functions of human skin are one of the research focuses. Tactile sensors can generate corresponding electrical signals in response to external stress stimuli and are widely used in artificial intelligence, human-computer interaction, bioinformatics detection and other fields. A deep understanding of the perception principle of human skin is an important prerequisite for designing bionic tactile sensors. It is understood that human skin is a very "powerful" tactile sensor that can detect the intensity and pattern of various stimuli at the same time, and can distinguish between pressing, tapping and bending. This is mainly attributed to the four mechanical receptors (SA-I, II and FA-I, II) distributed in different areas of human skin. The mechanical receptors receive external stimuli and convert them into electronic signals. The combined signals of these four receptors are analyzed by the brain to obtain information such as the size, shape and texture of the object.