- Asynchronously Coded Electronic Skin can sense pressure, temperature, and humidity with ultra-high precision and responsiveness.
- It can detect touch 1,000 times faster than the human sensory nervous system.
- It can be integrated with artificial intelligence to enable high-performance machine-brain interfaces.
For years, scientists have been trying to equip machines with a sense of touch. They have equipped intelligent human-like androids and prosthetic with electronic skins, which enable skins to work naturally and collaboratively with humans to manipulate various structures in the environments.
These electronic skins consist of a large array of sensors to provide quick somatosensory perception. However, they transmit tactile data from sensors serially, which results in higher readout latency bottlenecks.
Now, researchers at the University of Singapore have developed the Asynchronously Coded Electronic Skin (ACES) with ultra-high responsiveness and robustness. Inspired by the nervous system, ACES can sense pressure, temperature, and humidity. It can be paired with almost all types of sensor skin layers to operate effectively as an electronic skin.
How It Works?
The human body contains nearly 72 kilometers of nerves connecting the brain, skin, muscles. In this study, researchers used the human nervous system as inspiration to develop an electronic skin for robots.
The skin (ACES) is made of silicon layers covered with 249 sensors. It can simultaneously transmit thermo-tactile data while maintaining astonishingly low readout latencies, even with large numbers of sensors (up to 10,000).
Reference: ScienceMag | DOI:10.1126/scirobotics.aax2198 | University of Singapore
They demonstrated prototype arrays of up to 240 artificial mechanoreceptors. All were able to transmit data asynchronously with 1 millisecond of latency while maintaining a temporal precision of less than 60 nanoseconds.
Such low latency and ultra-high precision allow electronic skin to resolve fine spatiotemporal features essential for quick tactile perception.
Image credit: Mario Tama/Getty
This is a major breakthrough because for the first time researchers have been able to build an electronic skin that allows multiple sensors to feedback to one receiver, and act as a whole system instead of individual electrodes. Since sensors are connected in parallel, the whole system can still function normally even if the individual receptors are damaged.
There is one potential disadvantage of ACES – transmitter or receiver cannot tell whether a loss of sensing event has occurred. However, the architecture enables simple transmitter implementations and impressive timing precision of stimuli events, which are extremely important factors for electronic skin applications where real-time motions are critical.
Read: Artificial Intelligence Helps Robot Recognize Objects By Touch
The research team is currently working with neuroscientists and engineers to help restore a sense of touch to prosthetic hands. They believe their electronic skin could advance machine-human-environment interactions for autonomous anthropomorphic robots and could be integrated with artificial intelligence to enable high-performance machine-brain interfaces.
