soft lithography

This study addresses the challenge of inefficient fluid manipulation and mixing in microfluidic systems under laminar flow, which is inherently constrained by fixed channel boundaries. To overcome this limitation, the authors propose a programmable soft magnetic membrane actuator fabricated via template-assisted magnetization to encode predefined magnetic domain patterns. Upon exposure to an external magnetic field, the membrane undergoes controllable sinusoidal deformations, enabling dynamic reconfiguration of the microchannel walls. This approach achieves, for the first time, wireless and remote actuation of active microchannel shape adaptation, significantly enhancing mixing efficiency in laminar flow regimes. The platform offers a versatile foundation for developing next-generation lab-on-a-chip systems featuring deformable architectures and wireless control.

This work addresses the limitations of conventional acoustic microrobots, whose operational lifetime and resonance frequency stability are compromised by rapid dissolution of gas bubbles within sealed microcavities. To overcome this, the authors propose a soft acoustic microrobot encapsulated by a flexible polydimethylsiloxane (PDMS) membrane, which significantly suppresses gas diffusion and thereby enhances actuation stability. Integrated magnetic microparticles enable precise navigation under low-intensity magnetic fields. This design fundamentally mitigates bubble dissolution, allowing continuous and stable operation for over 24 hours even under high driving voltages. Furthermore, the platform supports scalable fabrication at the hundred-micrometer scale and enables controllable locomotion.

To address the limitations of weak actuation force, slow response, and poor environmental/biocompatibility in soft robotic ionic polymer actuators, this study presents a fully 3D-printed, biodegradable ionic membrane. Leveraging a synergistic direct ink writing (DIW) and fused deposition modeling (FDM) process, we achieve one-step fabrication with in situ encapsulation of a biodegradable ionic liquid between activated carbon/polymer composite layers—a first-of-its-kind approach. The resulting membrane exhibits exceptional structural integrity and dynamic performance: a bending curvature of 0.82 cm⁻¹ (the highest reported to date), maximum bending angle of 124°, blocking force of 0.76 mN, and operational frequency up to 2 Hz. This design uniquely integrates high electromechanical performance, environmental sustainability, and human biocompatibility, enabling rapid, customizable fabrication of soft actuators and human–machine interface devices.

Soft microrobots face a fundamental trade-off between the high brittleness of piezoceramic actuators (e.g., PZT) and the low bandwidth of polymeric actuators. Method: This work proposes a voltage-parallel multilayer piezoelectric actuator based on polyvinylidene fluoride (PVDF), integrating a thickness–layer-count co-optimized electromechanical performance model, thin-film stacking fabrication, and resonant driving. Contribution/Results: The design achieves efficient electromechanical transduction at low drive voltages while preserving flexibility: free deflection >3 mm, blocking force ≈20 mN, and operational bandwidth >500 Hz—effectively bridging the performance gap between rigid PZT and soft polymer actuators. Integrated into a planar microrobot, the actuator demonstrates robust disturbance-rejection locomotion, thereby advancing the design paradigm and application scope of soft microrobotic actuators.

Soft robotic systems face significant challenges in monolithically integrating actuation, sensing, and structural functionality within a single, scalable platform. Method: This work introduces the Monolithic Unit (MU) design paradigm—a unified 3D-printed architecture that co-integrates pneumatic actuation chambers, a compliant lattice-based load-bearing structure, and an embedded optical waveguide sensing network. Leveraging parametric modeling, lattice homogenization experiments, finite element analysis, and multi-objective optimization of waveguide routing, the MU enables concurrent design of mechanical performance and sensing sensitivity. Contribution/Results: Fabricated MUs span multiple scales and are successfully deployed in a two-finger soft gripper. They retain original actuation performance and structural stiffness while enabling high signal-to-noise-ratio, intrinsically conformal, real-time deformation sensing. This work presents the first scalable, fully monolithic integration of actuation, structure, and sensing—establishing a novel architectural foundation for embedded intelligence in soft robotics.

This work addresses the challenges of reproducibility and scalability in soft robot fabrication for rehabilitation and surgical applications. The authors propose a robust manufacturing process based on two-part silicone casting, integrating anti-clogging chamber design, hermetic sealing mechanisms, and an embedded thin-film flexible sensor integration method. System validation is achieved through finite element modeling, PID-based pneumatic control, and automated image processing. The protocol demonstrated high repeatability, with 24 successful fabrications performed independently by two operators. Actuator performance under step and sinusoidal inputs exhibited consistent response and controllable hysteresis, while simulated and experimental bending angles showed strong agreement, confirming the robustness and efficacy of the proposed methodology.

This work addresses the challenge of achieving fine-grained, distributed shape control in conventional soft robots by proposing a unit-level co-design approach. For the first time, bidirectional bellows actuators and curved beam structures are integrated within individual lattice units, enabling spatially selective deformation through embedded pneumatic actuation. The design facilitates heterogeneous unit assembly and selective pressure control, allowing diverse global deformation patterns to be generated without hardware reconfiguration. The method’s scalability, cyclic stability, and morphological versatility are demonstrated in a 3×3×3 array, which successfully executes bending, directional grasping, and asymmetric crawling via active-passive coupled actuation.

This work addresses the challenge of scaling microfluidic fuel cells to practically relevant power levels, which has been hindered by the high computational cost and poor scalability of traditional computational fluid dynamics (CFD) simulations. The authors propose a reduced-order modeling approach that accurately captures single-cell behavior while enabling efficient extension to large-scale stacks. This method dramatically reduces simulation time yet maintains excellent agreement with high-fidelity CFD results. For the first time, it enables efficient simulation and design support for macroscale microfluidic fuel cell systems with real-world applicability, overcoming the system-level scalability bottleneck inherent in conventional CFD and offering a viable pathway toward scalable microfluidic fuel cell design.

This study addresses the lack of high-precision, distinguishable bidirectional deformation sensing in soft lattice robots undergoing compression and extension. The authors propose BOAT, a dual-patterned optical waveguide sensor based on ellipsoidal geometric arrangements, co-printed with pneumatic artificial muscle–actuated lattice structures. By monitoring intensity changes in light transmission induced by waveguide bending, the system enables real-time identification of bidirectional deformation states. This work presents the first demonstration of monolithic co-printing of soft lattices integrated with bidirectional optical sensors, facilitating the development of synchronized digital twin systems. Experimental results show that under cyclic pressure loading of ±50 kPa, the system reliably distinguishes between tensile and compressive states with high repeatability, successfully realizing a physically-virtually synchronized digital shadow.

This work proposes a scalable micro-patterning fabrication approach based on cooperative multi-robot systems to overcome the limitations of conventional techniques, which are hindered by costly equipment and inefficient raster scanning. By integrating distributed multi-robot coordination with ergodic control, the method models the coverage task as a distribution-matching problem and enables efficient communication and task allocation through compressed trajectory histories. This novel framework breaks through traditional manufacturing bottlenecks, demonstrating in both simulations and physical experiments the ability to produce complex micro-patterns at scale with high efficiency. The resulting patterned metal surfaces exhibit significantly reduced friction coefficients, confirming the feasibility and effectiveness of the proposed approach.