Qing Shi

Single-cell microgels, engineered to replicate complex 3-D in vivo niches, have shown tremendous potential for advancing biomedical research. Studies have demonstrated that microgels recapitulating tissues’ stiffness facilitate the observation of cellular behavior in vitro. However, encapsulating single cells in microgels with composable stiffness gradients to understand cellular responses at microscale interfaces remains challenging. Here, we propose a versatile fabrication method that integrates an enhanced vision model with time-discrete bioprinting to encapsulate single cells in microgels featuring composable stiffness. An enhanced deep learning algorithm was developed to localize single cells in microscopic images, integrating traditional image processing algorithms for improved precision. By combining single-cell positioning data with a time-discrete array based on a digital micromirror device, dynamic digital masks can be generated to control each micromirror in real time. Using these masks, MDA-MB-231 cells were encapsulated in Gelatin methacrylate microgels with tunable stiffness ranging from 1.42 to 11.42 kPa. Experimental results indicated that when the microgel stiffness gradient exceeded 2 kPa, MDA-MB-231 cells exhibited a significant response, extending toward the softer regions of the microgels. These findings illustrate that our approach provides a valuable biofabrication strategy for investigating single-cell behaviors in microgels, showing great potential for applications in 3-D single-cell research.

Quadrupeds in nature achieve agile locomotion through a rhythmic flexing of the spine in coordination with leg movement. This dynamic synergy enhances their speed and stability, reflecting a kind of physical intelligence encoded in their bodies. However, identifying this spine-leg synergy and embodying it in robotics to enhance locomotor propulsion remains challenging. To address this, we developed FLEXOR (fast legged robot with a flexible spine for optimal running), which uses dual-joint coupled spine and elastic legs to capture dynamic spine-leg synergy in rapid propulsion of small-scale quadrupeds. Through simulation and physical experiments of FLEXOR, we identified an optimal dynamic spine-leg synergy that maximizes locomotor capabilities by aligning the ground reaction force (GRF) for effective forward propulsion. Moreover, the dual-joint coupled spine amplifies the actuator torque, thereby increasing both the GRF magnitude and the propulsive output without additional energy input. By harnessing this synergy and its structural advantage, FLEXOR achieves a substantial increase in speed with a reduced cost of transport, outperforming state-of-the-art quadruped robots with flexible spines. Through an extended dynamics framework and robophysical validation, we further demonstrate that the optimal spine-leg synergy generalizes robustly across diverse spine-leg morphologies and physical scales. This work broadens our understanding of the essence of synergistic locomotion in animals and is potentially applicable for the design and control of embodied intelligence–driven robots.

Agile and controllable motion is essential for active micromotors to navigate various microscale scenarios and manipulate the microscopic world, where parallel navigation and independent control are highly desirable. However, most micromotor navigation strategies suffer from limited agility and poor predictability, with motion typically confined to single-micromotor self-propulsion along predetermined trajectories. Herein, we propose an artificial intelligence (AI)–assisted optoelectronic control strategy that involves converting stochastic micromotor self-propulsion into controllable omnidirectional motion by synergistically exploiting multiple electrokinetic mechanisms. This strategy enables independent navigation of individual micromotors while simultaneously supporting parallel manipulation. By spatiotemporally configuring two or more optical patterns, agile motion primitives, such as directional propulsion, passive propulsion, in situ U-turns, and motion pause and restarting, were developed. To improve navigation robustness under coupled electrokinetic effects, a spatial-temporal AI model was developed for accurately predicting micromotor motion to facilitate the optimization of dynamic guidance schemes. These motion primitives are sequentially integrated and automatedly switched along long-term, reconfigurable trajectories, thereby enabling continuous navigation guided by discrete optical patterns. Independent control of active micromotors was demonstrated through the parallel manipulation of multiple Janus micromotors that were navigating intricate networks, in which each micromotor followed individual trajectories and adapted in real time to local terrain variations. This work showcased an agile and predictable navigation strategy for active micromotors that facilitates independent and massively parallel manipulation in intricate terrains, thus opening further possibilities for advanced applications.

Digital light processing (DLP) enables rapid fabrication of photocurable hydrogel microstructures, which serve as critical functional components in micro-optical systems and microrobotics, and whose performance depends on micron-scale morphological features and local stiffness. Traditional visual feedback methods provide horizontal data but struggle to quantify axial topography and local stiffness in real-time. Although digital holographic microscopy (DHM) offers advantages for sample characterization, its real-time capabilities are limited by the challenge of dynamically correcting for optical distortions. To address this, we present a real-time feedback control algorithm featuring a novel partial matrix Zernike fitting (PMZF) method. PMZF is analytically derived to estimate background phase distortion from partially occluded fields of view, enabling accurate phase reconstruction at 5 fps. With PMZF-based feedback, the system achieves 4.88 μm axial precision and 4.29 kPa stiffness precision, improving accuracy by 71.9% over open-loop DLP. Moreover, the spatially resolved control of stiffness within single microgels leads to region-specific fluorescent release, demonstrating a functional behavior not attainable with conventional printing. This work provides an effective closed-loop strategy for controlling both geometry and stiffness, paving the way for advanced functional devices in tissue engineering, MEMS, and beyond.

Flexible spines are critical to the remarkable agility and speed of animals. Translating this biological advantage to quadruped robots presents a significant control challenge, particularly in coordinating the spine and limbs for maximal velocity. In this work, we utilize reinforcement learning (RL) to develop high-speed locomotion for a bioinspired mouse robot with a lateral flexible spine. The resulting controller achieves motor performance that demonstrably surpasses non-spined and model-based methods. More importantly, our analysis reveals the principles behind this performance: the emergence of two distinct motion templates. For high-speed walking, the robot learns a “whip-like” spinal oscillation to increase leg swing frequency, while for agile turning, it adopts a dynamic “bend-and-straighten” pattern. These findings demonstrate the capability of RL to not only generate high-performance controllers but also to produce emergent strategies that, upon analysis, reveal underlying principles of high-speed, spine-driven locomotion.

Marine ecosystems, particularly coral reef communities, reveal how morphological diversification in fish species facilitates specialized locomotion through evolutionary optimization of body-fin coordination and hydrodynamic adaptations. Inspired by these biomechanical principles, we developed a morphology-encoded patterned magnetic millirobot (MPMR), whose anterior-to-posterior (AP) length ratio and body contour are predefined during fabrication to yield distinct hydrodynamic responses under the same uniform magnetic actuation. These MPMRs, with various morphologies, successfully emulated the undulatory swimming patterns of different fish species in a fluidic environment. Morphological differentiation in MPMRs has been shown to directly influence their motion performance, with an optimal AP ratio (approximately 1:1) and streamlined body contour maximizing propulsion efficiency. Furthermore, MPMRs with distinct morphologies display different frequency-dependent responses to magnetic actuation, leading to morphology-specific velocity profiles. By leveraging these morphology-encoded performance variations, we achieved effective selective control and multitarget delivery of multiple MPMRs under uniform magnetic fields, both in vitro and ex vivo (gastrointestinal tissue). These findings provide a foundation for future designs of flexible millirobots in similar environments and serve as a reference for advancing selective control methods for multiple millirobots in uniform magnetic fields.