Soft Control Interface for Highly Dexterous Unilateral Remote Palpation

Abstract

Achieving telepresence is the biggest challenge preventing healthcare robotics to permeate the field of primary care and performing tasks such as remote palpation. To do so, the user interface is of fundamental importance: ideally, it would reproduce similar conditions to the in-person task, resulting in intuitive and natural. Many studies focused on the high fidelity of the haptic and tactile feedback, overlooking the importance of the control interface. On the other hand, state- of-the-art soft control interfaces showcase high telepresence but low dexterity, limiting the user's range of motion. In this paper, a soft control interface for highly dexterous remote palpation, up to 5 degrees of freedom, is presented. The aim of this work is to use morphological computation to intrinsically encode hand trajectories, showcasing 3 different designs to test and optimize the interface's embodied intelligence. Then, the performance of the proposed interface is analyzed as a controller of a unilateral teleoperated system. The results show that the device is able to correctly encode the hand's motion and reproduce complex trajectories up to 5 DoF. Compared with keyboard control, it replicates 2D trajectories in a similar fashion, while maintaining the intuitiveness and naturalness of palpation.