Neuromorphic e-skin is fast, scalable and robust: Page 2 of 3

July 19, 2019 // By Julien Happich
A team of researchers from the National University of Singapore has demonstrated a novel type of signal architecture that could make robotic and prosthetic electronic skins much faster and more robust than today’s sequentially sampled tactile sensor arrays.

This signal architecture is particularly robust against damage, as all the sensors can simply be connected through the same conductive substrate, regardless of their distribution or shape. Even cuts or regular tear and wear won’t accidentally disconnect them as it would be the case with matrix-based addressing, and “killing several receptors” won’t affect the signal of others.

A multimodal ACES sensor array mimics the biological
sensory model, converting tactile stimuli into events that
are concurrently propagated.

The ACES platform also ensures individual receptors are plug and play, making it easy to combine multiple patches of ACES e-skins via a common conductive substrate, so they can be dynamically rearranged.

“This ACES signalling scheme is capable of propagating pulse signatures (events) from thousands of ACES receptors while preserving the relative time differences between the different tactile stimuli, which allows the rapid detection of object slippage, a key feature for grasp stability during in-hand object manipulation”, the authors wrote.

Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.