Search Results (293)

Semantic segmentation of workpieces and die cavities is critical for intelligent process monitoring and quality control in hammer die-forging. However, the field of 3D point cloud segmentation currently faces prominent limitations in forging scenario adaptation: existing state-of-the-art (SOTA) methods are predominantly optimized for road driving or indoor scenes, where targets have stable poses and regular surfaces. They lack dedicated designs for capturing the fine-grained deformation characteristics of forging workpieces and alleviating multi-scale feature misalignment caused by large pose variations—key pain points in forging segmentation. Consequently, these methods fail to balance segmentation accuracy and real-time efficiency required for practical forging applications. To address this gap, this paper proposes a novel semantic segmentation framework fusing 3D point cloud and bird’s-eye-view (BEV) representations for complex die-forging scenes. Specifically, a Star-based encoding module is designed in the BEV encoding stage to enhance capture of fine-grained workpiece deformation characteristics. A hierarchical feature-offset alignment mechanism is developed in decoding to alleviate multi-scale spatial and semantic misalignment, facilitating efficient cross-layer fusion. Additionally, a weighted adaptive fusion module enables complementary information interaction between point cloud and BEV modalities to improve precision.We evaluate the proposed method on our self-constructed simulated and real die-forging point cloud datasets. The results show that when trained solely on simulated data and tested directly in real-world scenarios, our method achieves an mIoU that surpasses RPVNet by $1.1\%$. After fine-tuning with a small amount of real data, the mIoU further improves by $5\%$, reaching optimal performance. Full article

In marine hydrodynamics, the core of the boundary element method (BEM) lies in the numerical calculation of the free-surface Green’s function. With the rise of artificial intelligence, using neural networks to fit Green’s function has become a new trend, yet most existing studies are confined to fitting Green’s function in infinite water depth. In this paper, a neural network fitting method for a finite-depth Green’s function is proposed. The classical Multilayer Perceptron (MLP) network and the emerging Kolmogorov–Arnold Network (KAN) are employed to conduct global and partition-based fitting experiments. Experiments indicate that the partition-based KAN fitting model achieves higher fitting accuracy, with most regions reaching 4D fitting precision. For large-scale data input, the average time for the model to calculate a single Green’s function value is 0.0868 microseconds, which is significantly faster than the 0.1120 s required by the traditional numerical integration method. These results demonstrate that the KAN can serve as an accurate and efficient model for finite-depth Green’s functions. The proposed KAN-based fitting method not only reduces the computational cost of numerical evaluation of Green’s functions but also maintains high prediction precision, providing an alternative approach to accelerate BEM calculations for floating body hydrodynamic analysis. Full article

Road defect detection is essential for timely road damage repair and traffic safety assurance. However, existing object detection algorithms suffer from insufficient accuracy in detecting small road surface defects and are prone to missed detections and false alarms under complex lighting and background conditions. To address these challenges, this study proposes SDC-YOLOv8, an improved YOLOv8-based algorithm for road defect detection that employs attention-enhanced feature learning and adaptive feature reconstruction. The model incorporates three key innovations: (1) an SPPF-LSKA module that integrates Fast Spatial Pyramid Pooling with Large Separable Kernel Attention to enhance multi-scale feature representation and irregular defect modeling capabilities; (2) DySample dynamic upsampling that replaces conventional interpolation methods for adaptive feature reconstruction with reduced computational cost; and (3) a Coordinate Attention module strategically inserted to improve spatial localization accuracy under complex conditions. Comprehensive experiments on a public pothole dataset demonstrate that SDC-YOLOv8 achieves 78.0% mAP@0.5, 81.0% Precision, and 70.7% Recall while maintaining real-time performance at 85 FPS. Compared to the baseline YOLOv8n model, the proposed method improves mAP@0.5 by 2.0 percentage points, Precision by 3.3 percentage points, and Recall by 1.8 percentage points, yielding an F1 score of 75.5%. These results demonstrate that SDC-YOLOv8 effectively enhances small-target detection accuracy while preserving real-time processing capability, offering a practical and efficient solution for intelligent road defect detection applications. Full article

Artificial intelligence (AI) is becoming more and more integrated into manufacturing processes, revolutionizing conventional production, like CNC (Computer Numerical Control) machining. This study analyzes how large language models (LLMs), exemplified by ChatGPT, behave when tasked with G-code optimization for improving surface quality and productivity of automotive metal parts, with emphasis on systematically documenting failure modes and limitations that emerge when general-purpose AI encounters specialized manufacturing domains. Even if software programming remains essential for highly regulated sectors, free AI tools will be increasingly used due to advantages like cost-effectiveness, adaptability, and continuous innovation. The condition is that there is sufficient technical expertise available in-house. The experiment carried out involved milling three identical parts using a Haas VF-3 SS CNC machine. The G-code was generated by SolidCam and was optimized using ChatGPT considering user-specified criteria. The aim was to improve the quality of the part’s surface, as well as increase productivity. The measurements were performed using an ISR C-300 Portable Surface Roughness Tester and Axiom Too 3D measuring equipment. The experiment revealed that while AI-generated code achieved a 37% reduction in cycle time (from 2.39 to 1.45 min) and significantly improved surface roughness (Ra—arithmetic mean deviation of the evaluated profile—decreased from 0.68 µm to 0.11 µm—an 84% improvement), it critically eliminated the pocket-milling operation, resulting in a non-conforming part. The AI optimization also removed essential safety features including tool length compensation (G43/H codes) and return-to-safe-position commands (G28), which required manual intervention to prevent tool breakage and part damage. Critical analysis revealed that ChatGPT failures stemmed from three factors: (1) token-minimization bias in LLM training leading to removal of the longest code block (31% of total code), (2) lack of semantic understanding of machining geometry, and (3) absence of manufacturing safety constraints in the AI model. This study demonstrates that current free AI tools like ChatGPT can identify optimization opportunities but lack the contextual understanding and manufacturing safety protocols necessary for autonomous CNC programming in production environments, highlighting both the potential, but also the limitation, of free AI software for CNC programming. Full article

Defects on large-scale structural surfaces can compromise integrity and pose safety hazards, highlighting the need for efficient automated inspection. UAVs provide a flexible and effective platform for such inspections, yet traditional vision-based methods often require high computational resources and show limited sensitivity to small defects, restricting practical UAV deployment. To address these challenges, we propose ESS-DETR, a lightweight and high-precision detection model designed for UAV-based surface inspection, built upon core modules: EMO-inspired lightweight backbone that integrates convolution and efficient attention mechanisms to reduce parameters; Scale-Decoupled Loss that adaptively balances targets of various sizes to enhance accuracy and robustness for small and irregular defect patterns frequently encountered in UAV imagery; and SPPELAN multi-scale fusion module that improves feature discrimination under complex reflections, shadows, and lighting variations typical of aerial inspection environments. Experimental results demonstrate that ESS-DETR reduces computational complexity from 103.4 to 60.5 GFLOPs and achieves a Precision of 0.837, Recall of 0.738, and mAP of 79, outperforming Faster R-CNN, RT-DETR, and YOLOv11, particularly for small-scale defects, confirming that ESS-DETR effectively balances accuracy, efficiency, and onboard deployability, providing a practical solution for intelligent UAV-based surface inspection. Full article

The advent of sixth-generation (6G) imposes stringent demands on wireless networks, while traditional 2D rigid reconfigurable intelligent surfaces (RISs) face bottlenecks in regulatory freedom and scenario adaptability. To address this, stacked intelligent metasurfaces (SIMs) have emerged. This paper presents a systematic review of SIM technology. It first elaborates on the SIM multi-layer stacked architecture and wave-domain signal-processing principles, which overcome the spatial constraints of conventional RISs. Then, it analyzes challenges, including beamforming and channel estimation for SIM, and explores its application prospects in key 6G scenarios such as integrated sensing and communication (ISAC), low earth orbit (LEO) satellite communication, semantic communication, and UAV communication, as well as future trends like integration with machine learning and nonlinear devices. Finally, it summarizes the open challenges in low-complexity design, modeling and optimization, and performance evaluation, aiming to provide insights to promote the large-scale adoption of SIM in next-generation wireless communications. Full article

Large language models (LLMs) have rapidly transformed artificial intelligence applications across industries, yet their integration into production systems has unveiled critical security vulnerabilities, chief among them prompt injection attacks. This comprehensive review synthesizes research from 2023 to 2025, analyzing 45 key sources, industry security reports, and documented real-world exploits. We examine the taxonomy of prompt injection techniques, including direct jailbreaking and indirect injection through external content. The rise of AI agent systems and the Model Context Protocol (MCP) has dramatically expanded attack surfaces, introducing vulnerabilities such as tool poisoning and credential theft. We document critical incidents including GitHub Copilot’s CVE-2025-53773 remote code execution vulnerability (CVSS 9.6) and ChatGPT’s Windows license key exposure. Research demonstrates that just five carefully crafted documents can manipulate AI responses 90% of the time through Retrieval-Augmented Generation (RAG) poisoning. We propose PALADIN, a defense-in-depth framework implementing five protective layers. This review provides actionable mitigation strategies based on OWASP Top 10 for LLM Applications 2025, identifies fundamental limitations including the stochastic nature problem and alignment paradox, and proposes research directions for architecturally secure AI systems. Our analysis reveals that prompt injection represents a fundamental architectural vulnerability requiring defense-in-depth approaches rather than singular solutions. Full article

Microalgae are an evolutionarily ancient and morphologically diverse group of photosynthetic eukaryotes, with taxonomic resolution complicated by environmentally driven phenotypic plasticity. This study merges deep learning and explainable artificial intelligence (XAI) to establish a transparent, reliable, and biologically meaningful framework for green microalgae (Chlorophyta) classification. Microscope images from three morphologically distinct algal species—Desmodesmus flavescens, Desmodesmus subspicatus, and Tetradesmus dimorphus representing the genera Desmodesmus and Tetradesmus within Chlorophyta—were analyzed using twelve convolutional neural networks, including EfficientNet-B0–B7, DenseNet201, NASNetLarge, Xception, and ResNet152V2. A curated dataset comprising 3624 microscopic images from three Chlorophyta species was used, split into training, validation, and test subsets. All models were trained using standardized preprocessing and data augmentation procedures, including grayscale conversion, CLAHE-based contrast enhancement, rotation, flipping, and brightness normalization. The model’s performance was assessed using accuracy and loss metrics on independent test datasets, while interpretability was evaluated through saliency maps and Gradient-weighted Class Activation Mapping (Grad-CAM) visualizations. ResNet152V2 achieved the highest overall performance among all evaluated architectures, outperforming EfficientNet variants, NASNetLarge, and Xception in terms of macro F1-score. Visualization analysis showed that both Grad-CAM and saliency mapping consistently highlighted biologically relevant regions—including cell walls, surface ornamentation, and colony structures—confirming that the models relied on taxonomically meaningful features rather than background artifacts. The findings indicate that the integration of deep learning and XAI can attain consistently high test accuracy for microalgal species, even with constrained datasets. This approach enables automated taxonomy and supports biodiversity monitoring, ecological assessment, biomass optimization, and biodiesel production by integrating interpretability with high predictive accuracy. Full article

The development of robust autonomous maritime systems is fundamentally constrained by the “data silo” problem, where valuable operational data from disparate fleets remain isolated due to privacy concerns, severely limiting the scalability of general-purpose navigation intelligence. To address this barrier, we propose a novel Federated Meta-Transfer Learning (FMTL) framework that enables collaborative evolution of unmanned surface vehicle (USV) swarms while preserving data privacy. Our hierarchical approach orchestrates three synergistic stages: (1) transfer learning pre-trains a universal “Sea-Sense” foundation model on large-scale maritime data to establish fundamental navigation priors; (2) federated learning enables decentralized fleets to collaboratively refine this model through encrypted gradient aggregation, forming a distributed cognitive network; (3) meta-learning allows for rapid personalization to individual vessel dynamics with minimal adaptation trials. Comprehensive simulations across heterogeneous fleet distributions demonstrate that our federated model achieves a 95.4% average success rate across diverse maritime scenarios, significantly outperforming isolated specialist models (63.9–73.1%), while enabling zero-shot performance of 78.5% and few-shot adaptation within 8–12 episodes on unseen tasks. This work establishes a scalable, privacy-preserving paradigm for collective maritime intelligence through swarm-based learning. Full article

Functional steel surfaces engineered through tailored micro-nano structures are increasingly vital for various applications such as high-performance aerospace components, energy conversion systems and defense equipment. Femtosecond laser filament processing is a recently proposed remote fabrication technique, showing the capability of fabricating micro-nano structures on irregular and large-area surfaces without the need of tight focusing. Nevertheless, the mechanisms underlying the formation of filament-induced structures remain not fully understood. Here we systematically investigate the formation mechanisms of filament-induced micro-nano structures on stainless steel surfaces by processing stainless steel in three manners: point, line, and area. We clarify the decisive role of the unique core–reservoir energy distribution of the filament in the formation of filament-induced micro-nano structures, and reveal that ablation, molten metal flow, and metal vapor condensation jointly drive the structure evolution through a dynamic interplay of competition and coupling, giving rise to the sequential morphological transitions of surface structures, from laser-induced periodic surface structures to ripple-like, crater-like, honeycomb-like, and ultimately taro-leaf-like structures. Our work not only clarifies the mechanisms of femtosecond laser filament processed morphological structures on steels but also provides insights onto intelligent manufacturing and design of advanced functional steel materials. Full article

To tackle the challenges faced by traditional wheeled tractors, whose steering systems have low flexibility and a large turning radius, and thus make turning hard in small fields and greenhouses, this paper proposes a differential steering control technology for battery-electric unmanned tractors. This innovative approach enables zero-radius turning while delivering environmental and economic advantages. Firstly, the system architecture and key components of the battery-electric unmanned tractor with differential steering are designed, including the mechanical structure, wheel-drive system, electrical system, and power battery. Based on the proposed system architecture, a multi-physics coupled model is established, covering the motor, reducer, battery, driver, vehicle body, and the relationship between tires and road surfaces. A multi-closed-loop control algorithm, regulating both the motor speed and yaw angular velocity of the tractor, is developed for differential steering control. The validation, conducted via a digital simulation platform, yields critical state curves for motor current, torque, speed, and vehicle rotation. This study establishes a novel theoretical framework for unmanned tractor control, with prototype development guided by the proposed methodology. Experimental validation of zero-radius steering confirms the efficacy of differential steering in battery-electric platforms. The research outcomes provide theoretical basis and technical references for advancing intelligent and electric agricultural equipment. Full article

The operational efficiency and precision of boom sprayers, as critical equipment for protecting field crops, are vital to global food security and agricultural sustainability. In precision agriculture systems, achieving uniform pesticide application fundamentally depends on maintaining stable boom posture during operation. However, severe boom vibration not only directly causes issues like missed spraying, double spraying, and pesticide drift but also represents a critical bottleneck constraining its functional realization in cutting-edge applications. Despite its importance, achieving absolute boom stability is a complex task. Its suspension system design faces a fundamental technical contradiction: effectively isolating high-frequency vehicle vibrations caused by ground surfaces while precisely following large-scale, low-frequency slope variations in the field. This paper systematically traces the evolutionary path of self-balancing boom technology in addressing this core contradiction. First, the paper conducts a dynamic analysis of the root causes of boom instability and the mechanism of its detrimental physical effects on spray quality. This serves as a foundation for the subsequent discussion on technical approaches for boom support and balancing systems. The paper also delves into the evolution of sensing technology, from “single-point height measurement” to “point cloud morphology perception,” and provides a detailed analysis of control strategies from classical PID to modern robust control and artificial intelligence methods. Furthermore, this paper explores the deep integration of this technology with precision agriculture applications, such as variable rate application and autonomous navigation. In conclusion, the paper summarizes the main challenges facing current technology and outlines future development trends, aiming to provide a comprehensive reference for research and development in this field. Full article

Smart vesicle therapeutics represent a transformative frontier in nanomedicine, offering precise, biocompatible, and adaptable platforms for drug delivery and theranostic applications. This review explores recent advances in the design and engineering of liposomes, niosomes, polymersomes, and extracellular vesicles (EVs), emphasizing their capacity to integrate therapeutic and diagnostic functions within a single nanoscale system. By tailoring vesicle size, composition, and surface chemistry, researchers have achieved improved pharmacokinetics, reduced immunogenicity, and fine-tuned control of drug release. Stimuli-responsive vesicles activated by pH, temperature, and redox gradients, or external fields enable spatiotemporal regulation of therapeutic action, while hybrid bio-inspired systems merge synthetic stability with natural targeting and biocompatibility. Theranostic vesicles further enhance precision medicine by allowing real-time imaging, monitoring, and adaptive control of treatment efficacy. Despite these advances, challenges in large-scale production, reproducibility, and regulatory standardization still limit clinical translation. Emerging solutions—such as microfluidic manufacturing, artificial intelligence-guided optimization, and multimodal imaging integration—are accelerating the development of personalized, high-performance vesicular therapeutics. Altogether, smart vesicle platforms exemplify the convergence of nanotechnology, biotechnology, and clinical science, driving the next generation of precision therapies that are safer, more effective, and tailored to individual patient needs. Full article

The rapid advancement of artificial technology is giving rise to new forms of cyber threats like memorization-based APT attacks, which not only pose significant risks to critical infrastructure but also present serious challenges to conventional security architectures. As a crucial service information system in railway passenger stations, the Railway Passenger Service System (RPSS) is particularly vulnerable due to its widespread terminal distribution and large attack surface. This paper focuses on two key challenges within the RPSS Cloud Center’s Double-Layer Dynamic Heterogeneous Redundancy (DDHR) architecture under such attacks: (i) the inability to accurately estimate redundant executor scheduling time, and (ii) the absence of an intelligent defense scheduling method capable of countering memorization-based attacks within a unified and quantifiable environment. To address these issues, we first establish the problem formulation of optimizing defender’s payoff under incomplete information, which applies information entropy of DDHR redundant executors to reflect attacking and defending behaviors. Then a method of estimating attacking time is proposed in order to overcome the difficulty in determining scheduling time due to incomplete information. Finally, we introduce the PPO_HE approach—a Proximal Policy Optimization (PPO) algorithm enhanced with quantifiable information Entropy and Heterogeneity of DDHR redundant executors. Extensive experiments were conducted for evaluation in terms of the two entropy-related metrics: information entropy decay amount and information entropy decay rate. Results demonstrate that the PPO_EH approach achieves the highest efficiency per scheduling operation in countering attacks and provides the longest resistance time against memorization-based attacks under identical initial information entropy conditions. Full article

As a fundamental and indispensable component of modern electronic devices, the printed circuit board (PCB) has a complex structure and highly integrated functions, with its manufacturing quality directly affecting the stability and reliability of electronic products. However, during large-scale automated PCB production, its surfaces are prone to various defects and imperfections due to uncontrollable factors, such as diverse manufacturing processes, stringent machining precision requirements, and complex production environments, which not only compromise product functionality but also pose potential safety hazards. At present, PCB defect detection in industry still predominantly relies on manual visual inspection, the efficiency and accuracy of which fall short of the automation and intelligence demands in modern electronics manufacturing. To address this issue, in this paper, we have made improvements based on the classical Faster-RCNN object detection framework. Firstly, ResNet-101 is employed to replace the conventional VGG-16 backbone, thereby enhancing the ability to perceive small objects and complex texture features. Then, we extract features from images by using deformable convolution in the backbone network to improve the model’s adaptive modeling capability for deformed objects and irregular defect regions. Finally, the Convolutional Block Attention Module is incorporated into the backbone, leveraging joint spatial and channel attention mechanisms to improve the effectiveness and discriminative power of feature representations. The experimental results demonstrate that the improved model achieves a 4.5% increase in mean average precision compared with the original Faster-RCNN. Moreover, the proposed method exhibits superior detection accuracy, robustness, and adaptability compared with mainstream object detection models, indicating strong potential for engineering applications and industrial deployment. Full article