Home // INTELLI 2021, The Tenth International Conference on Intelligent Systems and Applications // View article

Friction Invariant Object Reconstruction Using a Vibrissa-inspired Tactile Sensor Concept

Authors:
Lukas Merker
Joachim Steigenberger
Carsten Behn

Keywords: Vibrissa; tactile sensor; object scanning; object reconstruction; friction invariance

Abstract:
Vibrissae-inspired sensor concepts hold potential of complementing conventional sensors due to their collision robustness and increased near-field scanning range. In general, they consist of a slender elongated probe, one-sided attached to some kind of measuring device. Making contact with an object, the probe gets bent and transduces (mechanical) signals to its support/the measuring device. Modeling this process analytically, most approaches in literature are based on the assumption of an ideal contact (no stiction or friction). Within the present paper, we extend an existing tactile sensor model for object contour scanning and reconstruction taking first steps to implement frictional effects (Coulomb's law of friction). The probe is modeled as a rod, which is one-sided clamped. Shifting the clamping position relatively to the object of interest, represented by a plane and strictly convex object contour, the rod sweeps along the contour quasi-statically. This process is analyzed in two steps: (1) simulating scanning sweeps in order to generate the support reactions (observables) at the base of the rod theoretically, assuming the object contour and the coefficient of friction to be known, (2) using the generated support reactions from the previous step in order to reconstruct a sequence of contact points, finally approximating the original object contour. Our simulations suggest, that friction affects the observables but (surprisingly) not the actual reconstruction error. Finally, we present a novel approach for reconstructing friction parameters based on the support reactions.

Pages: 16 to 21

Copyright: Copyright (c) IARIA, 2021

Publication date: July 18, 2021

Published in: conference

ISSN: 2308-4065

ISBN: 978-1-61208-882-2

Location: Nice, France

Dates: from July 18, 2021 to July 22, 2021