Task space adaptation via the learning of gait controllers of magnetic soft millirobots

Learning Soft Millirobot Multimodal Locomotion with Sim-to-Real Transfer

With wireless multimodal locomotion capabilities, magnetic soft millirobots have emerged as potential minimally invasive medical robotic platforms. Due to their diverse shape programming capability, they can generate various locomotion modes, and their locomotion can be adapted to different environments by controlling the external magnetic field signal. Existing adaptation methods, however, are based on hand-tuned signals. Here, a learning-based adaptive magnetic soft millirobot multimodal locomotion framework empowered by sim-to-real transfer is presented. Developing a data-driven magnetic soft millirobot simulation environment, the periodic magnetic actuation signal is learned for a given soft millirobot in simulation. Then, the learned locomotion strategy is deployed to the real world using Bayesian optimization and Gaussian processes. Finally, automated domain recognition and locomotion adaptation for unknown environments using a Kullback-Leibler divergence-based probabilistic method are illustrated. This method can enable soft millirobot locomotion to quickly and continuously adapt to environmental changes and explore the actuation space for unanticipated solutions with minimum experimental cost.

Design and build of small-scale magnetic soft-bodied robots with multimodal locomotion

Small-scale magnetic soft-bodied robots can be designed to operate based on different locomotion modes to navigate and function inside unstructured, confined and varying environments. These soft millirobots may be useful for medical applications where the robots are tasked with moving inside the human body. Here we cover the entire process of developing small-scale magnetic soft-bodied millirobots with multimodal locomotion capability, including robot design, material preparation, robot fabrication, locomotion control and locomotion optimization. We describe in detail the design, fabrication and control of a sheet-shaped soft millirobot with 12 different locomotion modes for traversing different terrains, an ephyra jellyfish-inspired soft millirobot that can manipulate objects in liquids through various swimming modes, a larval zebrafish-inspired soft millirobot that can adjust its body stiffness for efficient propulsion in different swimming speeds and a dual stimuli-responsive sheet-shaped soft millirobot that can switch its locomotion modes automatically by responding to changes in the environmental temperature. The procedure is aimed at users with basic expertise in soft robot development. The procedure requires from a few days to several weeks to complete, depending on the degree of characterization required.

Bayesian optimization with unknown constraints in graphical skill models for compliant manipulation tasks using an industrial robot

This article focuses on learning manipulation skills from episodic reinforcement learning (RL) in unknown environments using industrial robot platforms. These platforms usually do not provide the required compliant control modalities to cope with unknown environments, e.g., force-sensitive contact tooling. This requires designing a suitable controller, while also providing the ability of adapting the controller parameters from collected evidence online. Thus, this work extends existing work on meta-learning for graphical skill-formalisms. First, we outline how a hybrid force–velocity controller can be applied to an industrial robot in order to design a graphical skill-formalism. This skill-formalism incorporates available task knowledge and allows for online episodic RL. In contrast to the existing work, we further propose to extend this skill-formalism by estimating the success probability of the task to be learned by means of factor graphs. This method allows assigning samples to individual factors, i.e., Gaussian processes (GPs) more efficiently and thus allows improving the learning performance, especially at early stages, where successful samples are usually only drawn in a sparse manner. Finally, we propose suitable constraint GP models and acquisition functions to obtain new samples in order to optimize the information gain, while also accounting for the success probability of the task. We outline a specific application example on the task of inserting the tip of a screwdriver into a screwhead with an industrial robot and evaluate our proposed extension against the state-of-the-art methods. The collected data outline that our method allows artificial agents to obtain feasible samples faster than existing approaches, while achieving a smaller regret value. This highlights the potential of our proposed work for future robotic applications.

Multicomponent and multifunctional integrated miniature soft robots

Miniature soft robots with elaborate structures and programmable physical properties could conduct micromanipulation with high precision as well as access confined and tortuous spaces, which promise benefits in medical tasks and environmental monitoring. To improve the functionalities and adaptability of miniature soft robots, a variety of integrated design and fabrication strategies have been proposed for the development of miniaturized soft robotic systems integrated with multicomponents and multifunctionalities. Combining the latest advancement in fabrication technologies, intelligent materials and active control methods enable these integrated robotic systems to adapt to increasingly complex application scenarios including precision medicine, intelligent electronics, and environmental and proprioceptive sensing. Herein, this review delivers an overview of various integration strategies applicable for miniature soft robotic systems, including semiconductor and microelectronic techniques, modular assembly based on self-healing and welding, modular assembly based on bonding agents, laser machining techniques, template assisted methods with modular material design, and 3D printing techniques. Emerging applications of the integrated miniature soft robots and perspectives for the future design of small-scale intelligent robots are discussed.