An ingestible, battery-free, tissue-adhering robotic interface for non-invasive and chronic electrostimulation of the gut.

Ingestible electronics have the capacity to transform our ability to effectively diagnose and potentially treat a broad set of conditions. Current applications could be significantly enhanced by addressing poor electrode-tissue contact, lack of navigation, short dwell time, and limited battery life. Here we report the development of an ingestible, battery-free, and tissue-adhering robotic interface (IngRI) for non-invasive and chronic electrostimulation of the gut, which addresses challenges associated with contact, navigation, retention, and powering (C-N-R-P) faced by existing ingestibles. We show that near-field inductive coupling operating near 13.56 MHz was sufficient to power and modulate the IngRI to deliver therapeutically relevant electrostimulation, which can be further enhanced by a bio-inspired, hydrogel-enabled adhesive interface. In swine models, we demonstrated the electrical interaction of IngRI with the gastric mucosa by recording conductive signaling from the subcutaneous space. We further observed changes in plasma ghrelin levels, the "hunger hormone," while IngRI was activated in vivo, demonstrating its clinical potential in regulating appetite and treating other endocrine conditions. The results of this study suggest that concepts inspired by soft and wireless skin-interfacing electronic devices can be applied to ingestible electronics with potential clinical applications for evaluating and treating gastrointestinal conditions.

Robust Learning-Based Control for Uncertain Nonlinear Systems With Validation on a Soft Robot.

Existing modeling and control methods for real-world systems typically deal with uncertainty and nonlinearity on a case-by-case basis. We present a universal and robust control framework for the general class of uncertain nonlinear systems. Our data-driven deep stochastic Koopman operator (DeSKO) model and robust learning control framework guarantee robust stability. DeSKO learns the uncertainty of dynamical systems by inferring a distribution of observables. The inferred distribution is used in our robust and stabilizing closed-loop controller for dynamical systems. We also develop a model predictive control framework with integral action to compensate for run-time parametric uncertainty, such as manipulating unknown objects. Modeling and control experiments in simulation show that our presented framework is more robust and scalable for robotic systems than state-of-the-art controllers using deep Koopman operators and reinforcement learning (RL) methods. We demonstrate that our method resists previously unseen uncertainties, such as external disturbances, at a magnitude of up to five times the maximum control input. Furthermore, we test our DeSKO-based control framework on a real-world soft robotic arm. It shows that our framework outperforms model-based controllers that have full knowledge of the model parameters, and the controller can conduct object pick-and-place tasks without further training. Our approach opens up new possibilities in robustly managing internal or external uncertainty while controlling high-dimensional nonlinear systems in a learning framework. This approach serves as a foundation to greatly simplify high-level control and decision-making for robots.