Skin-Inspired Tactile Intelligence for Robots by Spike Train-based Artificial Skin

Abstract

This research project aims to advance a skin-inspired tactile sensing system which incorporates artificial receptors in a 3D layout and self-generates spike train signals containing position-encoded information. The sensing system is designed to wrap up the entire surface of robot fingers and to require only a limited number of addressing lines even though the density of the artificial receptors is comparable to that of the mechanoreceptors in human fingers. A computational spiking neural network (SNN) model is developed to decode the combination of the spike train signals generated from the different sets of artificial receptors. The effects of periodic versus random distributions of the artificial receptors on tactile perception is investigated to find optimal sensor distributions. The tactile intelligence algorithm will be applied to a robotic perception-action feedback loop to recognize random objects without vision.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 16, 2024
Source ID
FA23862314029

Entities

People

  • Unyong Jeong

Organizations

  • Air Force Office of Scientific Research
  • Pohang University of Science and Technology
  • United States Air Force

Tags

Readers

  • Neural Network Machine Learning.
  • Neuroscience
  • Robotics and Automation.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • AI & ML - DoD AI Strategy
  • AI & ML - Neural Networks
  • Autonomy