Tactile sensing on deformed surfaces with electrical impedance tomography

Electrical Impedance Tomography (EIT)-based tactile sensors are emerging as a promising solution in robotic sensing, attributed to their cost-effectiveness, safety, and scalability due to their sparse electrode configurations. Capable of covering extensive areas of the robots' surface, these sensors showcase great potential for widespread applications. However, the pronounced sensitivity of EIT-based tactile sensors to deformations remains a persistent challenge when applied to soft systems. This paper presents a dual-modal tactile reconstruction approach that tracks and compensates for surface deformations during tactile sensing. This method first captures the complex deformations of the target object and then performs the reconstruction of tactile interactions with an updated EIT forward model, employing
$l$
1 regularisation. We first validate our method through numerical simulations, achieving a relative image error ranging from 0.6104 to 0.8106 and a correlation coefficient between 0.6076 and 0.8289. Further real-world evaluation is conducted using a hydrogel-based tactile sensor under a variety of deformation scenarios. The results affirm the effectiveness of the proposed approach, especially when integrated with a sensor system tailored for deformation sensing, paving the way for enhanced tactile interaction in soft and highly deformable robotic applications.