Encoding and Decoding of Static Information in Tactile Sensing Systems

Abstract

Human and robot tactile sensing can be accomplished by arrays of mechanosensors embedded in a deformable medium. When an object comes in contact with the surface of the medium, information about the shape of the medium's surface and the force distribution within the region of contact on the surface is available in the stress/strain states at the sensor locations within the medium. The mechanosensors transduce these signals, and the problem for the central processor is to reliably and efficiently infer the contact state and the object properties on the surface from the sensor signals. In this paper, a frequency-domain approach is used to solve the problem of encoding and decoding mechanosensory information. A Solution to the encoding problem is given with the medium modeled as a general three-dimensional infinite half-space composed of a linear elastic material and subjected to three-dimensional loads. It is shown that considerations of symmetry and bandwidth of sensor response uniquely determine the optimal stress/strain components the sensors need to transduce. It is further shown how the decoding leads to an ill-posed problem, and how that problem can be effeciently solved in the frequency domain using a regularized inverse such as the multivariate Wiener inverse. The results are the applied to the encoding and decoding of contact with a shaped object. It is shown that the solution can also be used in pseudodynamic problems, such as the estimation of the onset of slip.