Control the Soft Robot Arm with its Physical Twin

To exploit the compliant capabilities of soft robot arms we require controller which can exploit their physical capabilities. Teleoperation, leveraging a human in the loop, is a key step towards achieving more complex control strategies. Whilst teleoperation is widely used for rigid robots, for soft robots we require teleoperation methods where the configuration of the whole body is considered. We propose a method of using an identical 'physical twin', or demonstrator of the robot. This tendon robot can be back-driven, with the tendon lengths providing configuration perception, and enabling a direct mapping of tendon lengths for the execture. We demonstrate how this teleoperation across the entire configuration of the robot enables complex interactions with exploit the envrionment, such as squeezing into gaps. We also show how this method can generalize to robots which are a larger scale that the physical twin, and how, tuneability of the stiffness properties of the physical twin simplify its use.

Online Imitation Learning for Manipulation via Decaying Relative Correction through Teleoperation

Teleoperated robotic manipulators enable the collection of demonstration data, which can be used to train control policies through imitation learning. However, such methods can require significant amounts of training data to develop robust policies or adapt them to new and unseen tasks. While expert feedback can significantly enhance policy performance, providing continuous feedback can be cognitively demanding and time-consuming for experts. To address this challenge, we propose to use a cable-driven teleoperation system which can provide spatial corrections with 6 degree of freedom to the trajectories generated by a policy model. Specifically, we propose a correction method termed Decaying Relative Correction (DRC) which is based upon the spatial offset vector provided by the expert and exists temporarily, and which reduces the intervention steps required by an expert. Our results demonstrate that DRC reduces the required expert intervention rate by 30\% compared to a standard absolute corrective method. Furthermore, we show that integrating DRC within an online imitation learning framework rapidly increases the success rate of manipulation tasks such as raspberry harvesting and cloth wiping.