The efficacy of interaction behavior and internal stiffness control for embodied information gain in haptic perception

Abstract

Haptic perception in biological systems not only depends on the environmental conditions, but also on the behavioral state and the internal impedance of the embodiment because proprioceptive sensors are embedded in the muscle and tendons used for actuation. A simple example of such a phenomenon can be found when people are asked to palpate a soft tissue to identify a stiff-inclusion. People tend to perform a variety of palpation strategies depending on their previous knowledge and the desired information. Does this mean that the probing behavioral variables and internal muscle impedance parameters and their interaction with given environmental conditions play a role in the perception information gain during the estimation of soft tissue's properties? In this paper, we use a two-degree of freedom laboratory-made variable stiffness and indentation probe to investigate how the modulation of probing behavioral and internal stiffness variables can affect the accuracy of the depth estimation of stiff inclusions in artificial silicon phantom tissue using information gain metrics based on prior knowledge in form of memory primitives.