Reconstructing External Force on the Circumferential Body of Continuum Robot With Embedded Proprioceptive Sensors

The compliance of continuum robot poses difficulty in designing the controller, as the robot does not have position encoder and could be subject to uncertain external force (UEF). The former can be resolved through external sensors, while the latter issue still needs further investigation. This article presents a novel robot design that embeds eGaIn sensors into the soft material, and the elongation in the robot chamber can lead to different changes in resistance values. The sensors act as a proprioceptive mechanism to sense the presence of the UEF. To locate the UEF, the robot surface is divided into multiple regions, and the column and row positions with the force can be solved using Hidden Markov Model and regularization algorithm. Through the virtual work principle, the force magnitude can be also evaluated. Experimental results confirm that, with possible four columns and five rows, the accuracy of finding the actual column and row positions are, respectively, $ 97\%$ and $ 98.5\%$. The mean error in evaluating the tip position and the force magnitude is about $ \text{4}\;\text{mm}$ and $ \text{0.23}\;\text{N}$. This article offers a new approach to obtain the information of the UEF, allowing continuum robots to work in a constrained environment.