Biological organisms often have remarkable multifunctionality through intricate structures, such as concurrent shape morphing and stiffness variation in the octopus. Soft robots, which are inspired by natural creatures, usually require the integration of separate modules to achieve these various functions. As a result, the whole structure is cumbersome, and the control system is complex, often involving multiple control loops to finish a required task. Here, inspired by the scales that cover creatures like pangolins and fish, we developed a robotic structure that can vary its stiffness and change shape simultaneously in a highly integrated, compact body. The scale-inspired layered structure (SAILS) was enabled by the inversely designed programmable surface patterns of the scales. After fabrication, SAILS was inherently soft and flexible. When sealed in an elastic envelope and subjected to negative confining pressure, it transitioned to its designated shape and concurrently became stiff. SAILS could be actuated at frequencies as high as 5 hertz and achieved an apparent bending modulus change of up to 53 times between its soft and stiff states. We further demonstrated both the versatility of SAILS by developing a soft robot that is amphibious and adaptive and tunable landing systems for drones with the capacity to accommodate different loads.

Treatment of cancer often involves uses of multiple therapeutic strategies with different mechanisms of action. In this study we investigated combinations of nanosecond pulsed electric fields (nsPEF) with low concentrations of gemcitabine on human oral cancer cells. Cells (Cal-27) were treated with pulse parameters (20 pulses, 100 ns in duration, intensities of 10, 30 and 60 kV/cm) and then cultured in medium with 0.01 µg/ml gemcitabine. Proliferation, apoptosis/necrosis, invasion and morphology of those cells were examined using MTT, flow cytometry, clonogenics, transwell migration and TEM assay. Results show that combination treatments of gemcitabine and nsPEFs exhibited significant synergistic activities versus individual treatments for inhibiting oral cancer cell proliferation and inducing apoptosis and necrosis. However, there was no apparent synergism for cell invasion. By this we demonstrated synergistic inhibition of Cal-27 cells in vitro by nsPEFs and gemcitabine. Synergistic behavior indicates that these two treatments have different sites of action and combination treatment allows reduced doses of gemcitabine and lower nsPEF conditions, which may provide better treatment for patients than either treatment alone while reducing systemic toxicities.

The demand for advanced human-machine interfaces (HMIs) highlights the need for accurate measurement of muscle contraction states. Traditional methods, such as electromyography, cannot measure passive muscle contraction states, while optical and ultrasonic techniques suffer from motion artifacts due to their rigid transducers. To overcome these limitations, we developed a flexible multichannel electrical impedance sensor (FMEIS) for noninvasive detection of skeletal muscle contractions. By applying an imperceptible current, the FMEIS can target multiple deep muscles by capturing electric-field ripples generated by their contractions. With an ultrathin profile (~220 micrometers), a low elastic modulus (212.8 kilopascals) closely matching human skin, and engineered adhesive sensor surfaces, the FMEIS conforms nicely to human skin with minimized motion artifacts. The FMEIS achieved high accuracy in both hand gesture recognition and muscle force prediction using machine learning models. With demonstrated performance across multiple HMI applications, including human-robot collaboration, exoskeleton control, and virtual surgery, FMEIS shows great potential for future real-time collaborative HMI systems.

For bionic underwater robots, it is a great challenge for depth control due to model uncertainty and strong nonlinearity. To this end, we propose a data-driven model predictive control (MPC) approach using reinforcement learning (RL) for robotic penguin depth control. First, by imitating the underwater mode of the biological penguin, a robotic prototype with a tendon-driven head, two-degrees-of-freedom wings, and a tendon-driven tail was designed. Then, a data-driven MPC framework is proposed considering the structure and motion properties of the robotic penguin. Especially, a data-based learning environment is constructed using a motion capture system, computational fluid dynamics, and a backpropagation neural network. Meanwhile, to maximize the benefits of the controller while ensuring safety and stability, a data-driven MPC using the RL scheme is applied to approximate the optimal policy. Combined with an appropriate reward design and periodic training, the closed-loop controller performance is significantly improved, and the validity of the proposed framework is finally tested by extensive simulations and experiments. Notably, this work will provide valuable insights into the learning-based motion control of bionic underwater robots.

Model-based robot controllers require customized control-oriented models, involving expert knowledge and trial and error. Remarkably, the Koopman operator enables the control-oriented model identification through the input-output mapping set, breaking through the barriers of the customization services. However, in recent years, research on Koopman-based robot control has mostly focused on lifting function construction, deviating from the original intention of improving the controller performance. Thus, we propose a robot controller auto-generation framework using the Bayesian-based Koopman operator, significantly releasing labor and eliminating the design obstacle. First, we introduce the Koopman-based system identification method, and offer the basic lifting function design criteria. Then, a Bayesian-based optimization strategy with resource allocation is designed, which allows for the simultaneous optimization of the lifting function and controller. Next, taking model predictive control (MPC) as an example, a mission-oriented controller auto-generation framework is developed. Simulation and experimental results indicate that, under various robots and data sources, the proposed framework can effectively generate the robot controllers, and perform with a far greater level of mission accuracy than the un-optimized Koopman-based MPC. Meanwhile, the proposed technique exhibits an obvious compensation effect against disturbances, demonstrating its practicability in robot control.