Human Inspired Multi-Modal Robot Touch

  • Nicholas J Pestell

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)

Abstract

Humans have exquisite capabilities for dexterous manipulation, demonstrated by our ability to perform complex tasks such as surgical operations, playing the guitar or even peeling an orange. Robots have to progress significantly in order to match human performance in such tasks. It is clear that humans rely heavily on a sense of touch for dexterous manipulation, as is evident from trying to perform even the most simple of task with cold hands. Therefore, this thesis proposes that: (i) touch is an essential sensory modality for robots and (ii) human touch is an ideal model on which to base robot tactile perceptual systems. This thesis demonstrates how aspects of human touch, namely multi-modality, can be applied to tactile sensing in robots. Within a set of artificial sensory channels based on peripheral afferents and their associated encoding for discriminative touch are developed and applied to texture perception. Finally, biomimetic tactile sensors are applied to robot hands to aid grasping.
The success of these techniques provides evidence that human-inspiration is a fruitful approach to endowing robots with a sense of touch and is a concept which should be progressed into the natural extensions of this work.
Date of Award21 Jan 2021
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorNathan F Lepora (Supervisor)

Cite this

'