Learning Dynamic Tasks on a Large-scale Soft Robot in a Handful of Trials

Soft robots offer more flexibility, compliance, and adaptability than traditional rigid robots. They are also typically lighter and cheaper to manufacture. However, their use in real-world applications is limited due to modeling challenges and difficulties in integrating effective proprioceptive sensors. Large-scale soft robots ($\approx$ two meters in length) have greater modeling complexity due to increased inertia and related effects of gravity. Common efforts to ease these modeling difficulties such as assuming simple kinematic and dynamics models also limit the general capabilities of soft robots and are not applicable in tasks requiring fast, dynamic motion like throwing and hammering. To overcome these challenges, we propose a data-efficient Bayesian optimization-based approach for learning control policies for dynamic tasks on a large-scale soft robot. Our approach optimizes the task objective function directly from commanded pressures, without requiring approximate kinematics or dynamics as an intermediate step. We demonstrate the effectiveness of our approach through both simulated and real-world experiments.

Baloo: A Large-Scale Hybrid Soft Robotic Torso for Whole-Arm Manipulation

Soft robotic actuators and their inherent compliance can simplify the design of controllers when operating in contact-rich environments. With such structures we can accomplish high-impact, dynamic, and contact-rich tasks that would be difficult using conventional rigid robots which might either break the robot or the object without careful modeling and design of high bandwidth controllers. In order to explore the benefits of structural passive compliance and exploit them effectively, we present a prototype robotic torso named Baloo, designed with a hybrid rigid-soft methodology, incorporating both adaptability from soft components and strength from rigid components. Baloo consists of two meter-long, pneumatically-driven soft robot arms mounted on a rigid torso and driven vertically by a linear actuator. We explore some challenges inherent in controlling this type of robot and build on previous work with rigid robots to develop a joint-level neural-network adaptive controller to enable high performance tracking of highly nonlinear, time-varying soft robot dynamics. We also demonstrate a promising use case for the platform with several hardware experiments performing whole-body manipulation with large, heavy, and unwieldy objects. A video of our results can be viewed at https://youtu.be/eTUvBEVGKXY.