ECT Sensor for Symmetrical Robot Gripper Application

Abstract

Electrical capacitance tomography (ECT) is an imaging technique for visualizing permittivity distributions in industrial processes. This paper first time explores application of ECT in robotic grippers. In this case, the electrode configuration significantly affects imaging quality. If a group of 4 electrodes is mounted on each finger of a three-finger symmetrical gripper, the gap between groups of electrodes will affect image reconstruction, and the sensitivity matrix should change accordingly. This study investigates the inter-group electrode gap to intra-group electrode gap ratio (GGR) from 1:1 to 25:1 by software through simulation and experiment. Image reconstruction was conducted using linear back projection (LBP) and Landweber iterations, incorporating weighted electrode pairs, multi-angle fusion and subsequent optimization. Both simulation and experiment confirm the effectiveness of optimized imaging strategies. For stratified and annular distributions, multi-angle and optimized fusion consistently achieved high correlation coefficients for all GGRs, exceeding 0.90/0.77 and 0.85/0.91, respectively. For two-object distributions, optimized fusion maintained a correlation coefficient above 0.80 within the GGR range of 2:1 to 14:1. Experimental results at GGR = 5 demonstrated the effectiveness of the proposed methods, achieving high correlation coefficients of 0.91/0.96 for stratified targets, 0.94/0.87 for annular targets, and 0.86/0.82 for two-object targets, which closely aligns with simulation results. These results demonstrate that optimizing GGR and incorporating weighted electrode pairs can significantly enhance ECT image quality. The proposed methodology establishes a foundation for integrating ECT into robotic grippers, enabling reliable internal structural characterization and analysis.