Dynamic Error Reduction in Soft Tactile Sensors by Optimum Material Selection

Emerging applications in the field of robotics require tactile sensors with low dynamic errors. Such sensors are often made of elastomers, which exhibit time-dependent drift and hysteresis, leading to lower sensing accuracy. In this work, we systematically study the dominant material-inherent sources for the dynamic error within a tactile sensor, based upon a diverse range of commercially available elastomers. We demonstrate that the initial error associated with the Mullins effect can be completely eliminated, that the residual transient force error related to viscoelastic drift is reduced by 57%, and that the hysteresis force error is reduced by 20% with an optimized selection of suitable elastomers.