Design of TCP-actuator-driven, soft-tendon-integrated anthropomorphic dexterous hand: SoroAgilHand-1

This paper presents the design, analysis, and development of a dexterous hand driven by TCP actuators, termed the SoroAgilHand-1. Using the elastic stress of the TCP array with the measured Young's modulus, a mapping model of the tendon-contracted length to the joint angle of the dexterous hand was developed and validated by the recorded trajectory of the thumb-index finger. We introduced a self-sensing model for the TCP array in the SoroAgilHand-1 system and established an integrated closed-loop control system using the tensile sensors and strain-measuring devices. Considering the Kapandji test and the Feix taxonomy, the grasping performance of SoroAgilHand-1 was evaluated using postures from the collected data. The experimental results show that the soft dexterous hand can achieve 33 grasping postures and pass the Kapandji test, implying that it fulfils the anthropomorphic design goal using TCP actuators. Moreover, based on the grabbing stability test from the Feix taxonomy, the finger-tip trajectory of the closed-loop controller for the driven TCP was significantly improved and exhibited greater stability than that in open-loop driving.