ABSTRACT Millimeter‐scale robots operating at the water–air interface represent an emerging frontier in microsystems, particularly for tasks within confined and topologically complex aquatic environments. However, their development is hindered by scale‐dominated physical effects, the strong influence of surface tension and pressure resistance, which challenge both reliable structural fabrication and controllable multimodal locomotion. Here, we report a bioinspired design strategy that integrates tailored mechanical architectures, additive manufacturing, and stimuli‐responsive materials to create a Chlamydomonas ‐inspired millirobot (CI‐Robot). The CI‐Robot combines Marangoni propulsion with a flagellum‐mimetic capillary network and modular control elements, including photo‐responsive hydrogel valves and magnetic eyes. The gel‐valves gate fuel pathways on demand, enabling programmable switching between straight translation and rotational gaits driven by Marangoni surface‐tension gradients. Coupling valve‐defined outlet states with Marangoni‐enhanced mass transfer and capillary liquid retention further allows microsample capture and retention, including microplastics and bacteria (detection limit: 100 CFU/mL), with >30 min sample retention. In parallel, the magnetic‐eyes provide robust reorientation and guided navigation, enabling complex path planning, obstacle avoidance, and site‐specific payload release in confined, cluttered settings. By unifying programmable locomotion, environmental adaptability, and multifunctional execution within a single materials‐integrable platform, CI‐Robot provides a practical route toward precision sampling, responsive monitoring, and targeted intervention in confined aquatic systems.

Industrial printed circuit board (PCB) defect detection plays a vital role in ensuring the quality and reliability of electronic products. However, accurately identifying small defects in high-resolution PCB images remains a highly challenging task. To address this challenge, we propose MSAD-Net, a small-defect detection framework designed for PCB inspection. By integrating mask-driven learning with spatially adaptive downsampling, MSAD-Net significantly enhances the saliency and detection reliability of small targets. First, a mask-Gaussian pixel-level auxiliary supervision mechanism is introduced, aligning intermediate-layer target masks with anisotropic Gaussian distributions derived from bounding-box annotations, thereby improving the geometric representation accuracy of tiny objects. Second, a mask-guided attention module (MGAM) is developed to suppress background interference and substantially boost the foreground signal-to-noise ratio, further strengthening the spatial saliency of small targets. Finally, a spatially adaptive downsampling module (SADM) is constructed, which employs content-aware strategies to preserve semantic and edge details while reducing feature-map resolution. Extensive experiments on two public PCB defect datasets and a self-built PCBA component dataset demonstrate that MSAD-Net significantly outperforms mainstream detection models in small-defect detection. Moreover, it maintains a lightweight architecture while achieving efficient inference performance, validating its potential for high-precision PCB defect inspection.

High-speed and high-precision motion control remains a major challenge in Surface-Mount Technology (SMT), as conventional methods often fail to balance overshoot and settling time under varying task requirements. To address this challenge, a motion control architecture driven by task-specific performance indicators is developed. The proposed framework integrates adaptive position, velocity, and current loop algorithms with a multi-parameter cooperative regulation model, in which control parameters are dynamically updated via a bidirectional radial basis function neural network (BRBFNN) based iterative correction mechanism. Unlike approaches that optimize a single performance metric, the proposed dual-layer architecture enables real-time coordination between precision and response speed, allowing the control system to adapt to the heterogeneous demands of different components. Experimental validation was conducted on the Z-axis, recognized as the most demanding in terms of speed and precision. The results demonstrate that the method can effectively adjust control performance according to varying operational requirements. This study provides a feasible control solution for SMT equipment and offers a general framework for motion systems requiring simultaneous optimization of speed and precision.

Operational efficiency of placement machines constrains the overall production capacity of printed circuit board (PCB) assembly lines. Existing state-of-the-art algorithms face challenges, such as conflicts between multiple objectives and coupling within different problems. This article proposes a multiobjective hybrid evolutionary multitasking algorithm (MOHEMTA) to address PCB assembly optimization in beam-head placement machines. The algorithm divides the problem into pickup and placement tasks, leveraging implicit parallelism to enhance solution efficiency. A nozzle block encoding method and heuristic decoding strategies with domain knowledge are introduced to reduce encoding complexity and accelerate algorithm convergence. MOHEMTA enhances offspring population diversity and quality through an elitist strategy, evolutionary operators, and knowledge transfer mechanisms, while incorporating safeguards against negative transfer. Experiments demonstrate that the multiobjective solution performance and practical results of MOHEMTA are better than those of other state-of-the-art algorithms.

Metastasis, the leading cause of mortality in cancer patients, presents challenges for conventional photodynamic therapy (PDT) due to its reliance on localized light and oxygen application to tumors. To overcome these limitations, a self-sustained organelle-mimicking nanoreactor is developed here with programmable DNA switches that enables bio-chem-photocatalytic cascade-driven starvation-photodynamic synergistic therapy against tumor metastasis. Emulating the compartmentalization and positional assembly strategies found in living cells, this nano-organelle reactor allows quantitative co-compartmentalization of multiple functional modules for the designed self-illuminating chemiexcited PDT system. Within the space-confined nanoreactor, biofuel glucose is converted to hydrogen peroxide (H2O2) which enhances luminol-based chemiluminescence (CL), consequently driving the generation of photochemical singlet oxygen (1O2) via chemiluminescence resonance energy transfer. Meanwhile, hemoglobin functions as a synchronized oxygen supplier for both glucose oxidation and PDT, while also exhibiting peroxidase-like activity to produce hydroxyl radicals (·OH). Crucially, the nanoreactor keeps switching off in normal tissues, with on-demand activation in tumors through toehold-mediated strand displacement. These findings demonstrate that this nanoreactor, which is self-sufficient in light and oxygen and precise in striking tumors, presents a promising paradigm for managing highly metastatic cancers.

Metal oxide semiconductor (MOS) sensors have been broadly employed for gas detection. However, the distinctive chemical detection principle of MOS sensors renders them susceptible to interference by humidity. This paper proposes a novel method for suppressing humidity interference. This method possesses advantages, including rapid recognition, low power consumption, and the capability to achieve recognition using a single feature. Temperature modulation technology was used to record the sensor's resistance variation, and the response features of the obtained dynamic response signal are affected by both humidity and the measured gas. The suppression of humidity interference is achieved through the investigation of the response features of humidity and the measured gas. The response features of humidity and gas are amplified by column normalization. Principal Component Linear Discriminant Analysis (PC-LDA) is utilized to acquire humidity and gas information within data. Using ethanol as the primary experimental subject, suppression of humidity interference is achieved through a single feature (PC-LD2). Quantitative recognition of relative humidity (RH) can be achieved using PC-LD1. Multiple types of waveform and sensor were used to demonstrate the generalization of this method.

Research on memristive devices to seamlessly integrate and replicate the dynamic behaviors of biological synapses will illuminate the mechanisms underlying parallel processing and information storage in the human brain, thereby affording novel insights for the advancement of artificial intelligence. Here, an artificial electric synapse is demonstrated on a one-step Mo-selenized MoSe2 memristor, having not only long-term stable resistive switching characteristics (reset 0.51 ± 0.01 V, on/off ratio > 30, retention > 103 s) but also diverse electrically adjustable synaptic behaviors, including multilevel conductance (synaptic weight), excitatory postsynaptic current (EPSC), paired-pulse facilitation (PPF), long-term potentiation/depression (LTP/D), spike-timing-dependent plasticity (STDP), and especially activity-dependent synaptic plasticity (ADSP). More significantly, neuromorphic functions of both image edge extraction and biological perception imitation have been successfully achieved. These results present a promising design toward synaptic devices for advancing neuromorphic systems with integrated brain-like neural sensing, memory, and recognition.

With the rapid development of micro-robotics, non-mechanical stimulus-responsive water-air interface mini-robots have become a prominent focus in intelligent materials and environmentally responsive systems. However, their versatile application is challenged by a fundamental trade-off: simpler structures enable precise motion control, while complex configurations are often required for task execution, making it difficult to balance controllable locomotion with functional complexity. Inspired by Chlamydomonas, we have designed a water-air interface mini-robot with a sophisticated multifunctional architecture (CI-Robot), enabling both programmable motion and multifunctional execution, which demonstrated tremendous potential for application in confined aquatic environments and complex pipelines. The robot can achieve ultra-fast linear and rotational speeds (11.43 body/s, 8.98π rad/s), exceeding biological counterparts by 1.37- and 4.24-fold, via synergistic surface tension gradients and flagellar capillary mechanisms. The fluid-solid coupling simulation reveals the motion mechanism of CI-Robot in the transitional Reynolds regimes, in which the inertial force stabilizes the propulsion force, and the driving torque rapidly decreases to equilibrium (~15.21 μN, ~10⁻⁹ N·m), providing a theoretical basis for the analysis and regulation of the robot's motion behavior. The safe separation distance (~2/3 body length) without interference is determined by collective motion analysis, which guides the reasonable arrangement of CI-Robot group operation. Integrating propulsion and functional modules, the CI-Robot excels in obstacle avoidance, complex path planning, microplastic collection (up to 10 2 particles/mL), bacterial sampling (up to 100 CFU/mL) and site-specific molecular release, retaining samples for >30 minutes. This innovative mini-robot combining unparalleled speed, adaptability, and multifunctionality, will pave the way for transformative applications in cargo delivery, environmental monitoring, microplastic collection, and site-specific sampling in confined space.

In surface mount technology (SMT) assembly, the increasing complexity and miniaturization of electronic components pose critical challenges to balancing placement precision and production efficiency. This paper presents a quality-efficiency driven scheduling and process co-optimization (SPCO) methodology for the pick-and-place (PAP) process in SMT production lines. Leveraging a cyber-physical system framework integrated with automated optical inspection, the proposed approach dynamically couples offline scheduling with online process capability feedback to achieve adaptive allocation of components. A precision-aware SPCO model is formulated to assign components to placement heads based on real-time process capability indices, ensuring compliance with stringent precision constraints. To enable real-time deployment, a precision-prioritized allocation heuristic (PPAH) is introduced, supporting component-head assignments under heterogeneous head capabilities. Experiments on industrial datasets demonstrate that PPAH completely eliminates precision violations while improving the overall process capability margin by 2.6-fold compared to state-of-the-art benchmarks, with only a moderate increase in total PAP time. These results validate the effectiveness of the proposed co-optimization strategy in improving first-pass yield and robustness in high-mix SMT environments.

Time-optimal path generation is critical for maximizing throughput in high-speed PCB assembly, yet existing approaches predominantly focus on geometric distance minimization, overlooking the fundamental impact of acceleration dynamics and multi-axis coordination on temporal efficiency. This study addresses this gap by introducing a physics-based time estimator that explicitly models trapezoidal acceleration profiles for synchronized X/Y/Z/R-axis motions, enabling precise performance evaluation under realistic kinematic constraints. Experimental validation on production PCBs demonstrates that the proposed estimator achieves much higher estimation accuracy, outperforming conventional methods. When integrated as the objective function in multi-chromosome genetic algorithm optimization, time-optimal solutions effectively reduce actual movement times compared to distance-optimal baselines, despite requiring longer travel paths. These findings confirm the time-optimal estimator’s superiority over pure distance minimization, proving that peak efficiency is achieved by balancing travel distance with movement speed.

Detecting circulating tumor cells has exhibited great significance in treating cancers since its concentration is an index strongly associated with the development and transfer of the tumor. However, the present commercial method for CTC detection is still expensive, because special antibodies and complicated devices must be used for cell separation and imaging. Hence, it is quite necessary to apply alternative materials and methods to decrease the cost of CTC detection. In this article, we coated a cellulose acetate membrane with nanoparticles formed by the polymerization of melamine and furfural, creating a surface with nanoscale roughness for the highly efficient capture of the sparse CTCs in a blood sample. Subsequently, the CTCs on the surface can be quantitatively detected by colorimetry with the aid of a COF-based nanozyme. The detection limit (LOD) can be as low as 3 cells/mL, which is the lowest LOD among the colorimetric methods to our knowledge. Considering the low cost of fabricating the membrane for CTC capture and the robustness of nanozymes compared with natural enzymes, this CTC detection approach displays great potential to decrease the financial burden of commercial CTC detection.

Soft robots, inspired by living organisms in nature, are primarily made of soft materials, and can be used to perform delicate tasks due to their high flexibility, such as grasping and locomotion. However, it is a challenge to efficiently manufacture soft robots with complex functions. In recent years, 3D printing technology has greatly improved the efficiency and flexibility of manufacturing soft robots. Unlike traditional subtractive manufacturing technologies, 3D printing, as an additive manufacturing method, can directly produce parts of high quality and complex geometry for soft robots without manual errors or costly post-processing. In this review, we investigate the basic concepts and working principles of current 3D printing technologies, including stereolithography, selective laser sintering, material extrusion, and material jetting. The advantages and disadvantages of fabricating soft robots are discussed. Various 3D printing materials for soft robots are introduced, including elastomers, shape memory polymers, hydrogels, composites, and other materials. Their functions and limitations in soft robots are illustrated. The existing 3D-printed soft robots, including soft grippers, soft locomotion robots, and wearable soft robots, are demonstrated. Their application in industrial, manufacturing, service, and assistive medical fields is discussed. We summarize the challenges of 3D printing at the technical level, material level, and application level. The prospects of 3D printing technology in the field of soft robots are explored.

Bottlebrush polymers, macromolecules consisting of dense polymer side chains grafted from a central polymer backbone, have unique properties resulting from this well-defined molecular architecture. With the advent of controlled radical polymerization techniques, access to these architectures has become more readily available. However, synthetic challenges remain, including the need for intermediate purification, the use of toxic solvents, and challenges with achieving long bottlebrush architectures due to backbone entanglements. Herein, we report hybrid bonding bottlebrush polymers (systems integrating covalent and noncovalent bonding of structural units) consisting of poly(sodium 4-styrenesulfonate) (p(NaSS)) brushes grafted from a peptide amphiphile (PA) supramolecular polymer backbone. This was achieved using photoinitiated electron/energy transfer-reversible addition-fragmentation chain transfer (PET-RAFT) polymerization in water. The structure of the hybrid bonding bottlebrush architecture was characterized using cryogenic transmission electron microscopy, and its properties were probed using rheological measurements. We observed that hybrid bonding bottlebrush polymers were able to organize into block architectures containing domains with high brush grafting density and others with no observable brushes. This finding is possibly a result of dynamic behavior unique to supramolecular polymer backbones, enabling molecular exchange or translational diffusion of monomers along the length of the assemblies. The hybrid bottlebrush polymers exhibited higher solution viscosity at moderate shear, protected supramolecular polymer backbones from disassembly at high shear, and supported self-healing capabilities, depending on grafting densities. Our results demonstrate an opportunity for novel properties in easily synthesized bottlebrush polymer architectures built with supramolecular polymers that might be useful in biomedical applications or for aqueous lubrication.

Automated of gas and liquid classification technologies are of great in multiple fields including food production and human healthcare. Of these, fruit juice contains water, organic acids, minerals and other nutrients which offers a pleasant taste and promotes healthy condition. However, the main challenges faced by conventional components sensing technologies for juice classification are limited to the complexity of experimental preparation, bulky instrument, high consumption and susceptibility to contamination. Moisture Electricity Generation (MEG) technology has made it feasible to acquire energy from trace amounts of water or environmental humidity. This work proposes a novel sensing unit based on MEG technology. The unit mainly comprises non-woven fabric, hydroxylated carbon nanotubes, polyvinyl alcohol, a solution of sea salt and liquid alloy. By this approach, humid air (relative humidity 60%), pure water and juices from three fruits (lemon, kiwifruit, and clementine) have been successfully classified in 15 seconds. The classification accuracy can reach 90%. Electrical signals standard lines highlight the specific response between samples. The relative standard deviation of stable output section is 1.6% and the root-mean-square error between test data and the standard curve is less than 0.08, which indicates the stability, accuracy are fine. Besides, the sensing unit demonstrates an acceptable reusability. The presented approach may provide opportunities to improve sensing paradigms in industrial and medical settings.

The optimization of printed circuit board assembly (PCBA) for a beam head placement machine is a multivariable and multiconstraint combinatorial problem. Current techniques falter in solving a variety of PCBA problems since heuristic algorithms lack theoretical guarantees of optimality, and mathematical modeling methods have high computational complexity for the whole problem. This article proposes a novel two-phase optimization for PCBA, integrating the advantages of mathematical modeling with heuristic algorithms. We divide the problem into the head task assignment and the placement route schedule. For the former, an effective integer linear programming model with component partition is proposed, encompassing key efficiency-influencing factors. A recursive heuristic-based initial solution speeds up the solving convergence, while the reduction strategies enhance model solvability. For the placement route schedule, a tailored greedy algorithm yields high-quality solutions, leveraging the results of the model, and an aggregated route relink heuristic does further optimization. In addition, we propose a selection criterion for the solution pool of the model to pre-evaluate the placement movement, which builds the connection between the two phases. Finally, we validate the performance of the two-phase optimization, which provides an average efficiency improvement of 8.66%-21.83% compared to other mainstream research.

To solve the testing problem of ring surface, a non-splicing interference testing system is designed by combining digital phase measurement technology. Using closed-loop circuit feedback control to realize the precise control of PZT for small displacement, while collecting the N frame sequence interference light intensity map In (x, y). In order to reduce the adjustment error introduced by manual adjustment, the error correction matrix is derived based on the misalignment error model under the cylindrical coordinate system. It has been proved that the testing system can realize high precision detection of 360 degrees rotary cylindrical surface, and a high precision Talyrond565LT cylindrical degree instrument is used to measure the same sample for comparison validation. The experimental results show that the ring surface detection system studied in this paper can realize one-time and high-precision detection of the inner surface of 360 degrees cylinder, and the measurement accuracy PV is better than 0.4 lambda(0.25 mu m)and RMS is better than 0.2 lambda(0.12 mu m).

A highly efficient coupling of glycosyl stannanes and sulfonium salts enabled by synergistic Pd/Cu catalysis is disclosed, facilitating the construction of C-aryl/alkenyl glycals under mild conditions in high yields. The protocol tolerates a wide scope of functional groups including ketone, cyano, ester, amide, nitro, halide. The one-pot formal CH glycosylation starting from arene is demonstrated with a reaction sequence of dibenzothiophenylation/Stille coupling. Besides, a gram-scale reaction is performed successfully, showing the high applicability of this protocol.

提出一种基于 2D先验的 3D目标判定算法.首先用轻量级MobileNet网络替换经典SSD的VGG-16 网络,构建出MobileNet-SSD目标检测模型;其次,通过改进网络结构,提高模型对小目标的检测能力,并引入Focal Loss函数来解决正负样本不均衡和易分样本占比较高的问题;在相同数据集上,将改进算法与Faster R-CNN、YOLOv3 及MobileNet-SSD进行对比测试,其平均精度mAP分别提高了 7.2％、8.8％和 10.6％;最后,通过改进算法获取ROI,利用深度相机将二维ROI转换为ROI点云,并借助直通滤波来判断目标物体是否为真实场景物体,既省去了传统点云识别中的诸多步骤又避免了点云深度学习中三维数据集制作难度较大的问题,在识别速度和识别精度上达到了较好的平衡.

Accurate detection of weeds is a key technology for developing automated weeding equipment. To address the problems of high detection complexity and poor robustness resulting from the complex distribution and variety of weeds, we proposed a weed detection approach for vegetable seedling based on the improved YOLOv5 algorithm and image processing, implemented on a self-developed mobile robot platform. The weed detection complexity was reduced by indirectly detecting weeds through identifying vegetables, thus improving the detection accuracy and robustness. The convolutional block attention module（CBAM） attention module was added to the backbone feature extraction network of the YOLOv5 object detection algorithm to enhance the focus of the network on vegetable targets, and the Transformer module was added to enhance the global information capture capability. The results showed that the average detection accuracy of the improved YOLOv5 algorithm for vegetable targets could reach 95.7%, which was increased by 5.8%, 6.9%, 10.3%, 13.1%, 9.0%, 5.2%, and 3.2% compared with Faster R-CNN,SSD, EfficientDet, RetinaNet, YOLOv3, YOLOv4, and YOLOv5, respectively. The average detection time of the algorithm for a single run was 11 ms, indicating good real-time performance. The method defined green plants outside the vegetable border as weeds, and combined the extreme green（ExG） with the OTSU threshold segmentation method to segment weeds from the soil background. Finally, the weed connectivity domain was marked, followed by outputting the weed plasmids and detection frames. The proposed method could provide a technical reference for automated precision weeding in agriculture.