Printed circuit boards are versatile and highly printed, which can be widely used in various fields, and also provide new opportunities for the development of electronic information equipment. However, it is difficult to detect defects and faults during the production and use of printed circuit boards. In this paper, in view of the defects and fault detection of printed circuit boards, a deep learning detection method GSC YOLOv5 that integrates light-weight network and dual attention mechanism is proposed. First, GSC YOLOv5 is improved on the basis of YOLOv5. Ghost Conv and Ghost Bottleneck are used to realize the lightweight of the algorithm structure, reduce the number of parameters and floating point arithmetic (Flops) of the model. Second, the dual attention mechanism of Squeeze-and-Excitation Module (SE Module) and Convolutional Block Attention Module (CBAM) are introduced to optimize the performance of the algorithm, while improves the detection precision and real-time detection efficiency. Last, the experimental results show that compared with the original algorithm, GSC YOLOv5 reduces the amount of parameters and Flops by 50.38% and 55.52%, respectively, and compresses the model volume by 50.26%. Furthermore, the detection precision is increased by 2.41% and the recall rate is improved by 1.06%. At the same time, a real-times detection performance of $89.4~FPS$ is achieved, it improves by 65.18%. Therefore, the proposed algorithm is not only lightweight but also can achieve better performance, it can satisfy the detection requirements of printed circuit board.

With the continued advancement of Artificial Intelligence (AI) technology in Customized Manufacturing (CM), the current intelligence model, which separates ‘perception’ from ‘execution,’ lacks adaptability and generalizability. Embodied Intelligence (EI), an emerging technology emphasizing real-time environmental interaction and feedback, is expected to enable integrated intelligent manufacturing systems characterized by ‘perception-cognition-execution-feedback,’ enhancing the performance and intelligence of multifactor systems like the Internet of Manufacturing Things (IoMT), equipment management, and resource scheduling. However, current AI systems in CM remain isolated and lack methods for environmental interaction with multi-source perception and feedback, hindering the development of autonomous, evolving intelligent systems. To address the environmental interaction bottlenecks among diverse production elements in CM, this paper proposes a Circular Embodied Intelligence Manufacturing (CEIM) architecture aimed at enabling the fusion of heterogeneous production-element information. The objective is to enhance environmental perception and establish autonomous system evolution mechanisms, thereby optimizing production decision-making. By integrating sensor data and AI model outputs, and leveraging the advanced reasoning capabilities of large language models (LLMs), CEIM facilitates the semantic fusion of multiple production elements—including IoMT, intelligent devices, and manufacturing resources—to enable the deployment of EI applications in CM. The implementation of CEIM is illustrated and validated through a case study on a customized gift box packaging platform. Finally, this paper discusses the opportunities and challenges of applying EI in manufacturing and aims to provide insights into the future development of EI-oriented manufacturing systems.

In this paper, a deep learning-based method is proposed for the fast and high-precision detection and digital reconstruction of damaged areas in grotto murals. First, the YOLO Mural breakage detection algorithm is presented, which is based on the YOLOv10 algorithm. The algorithm enhances the backbone network structure through the incorporation of a Star Block module and the parameter-free attention mechanism, SimAM. This modification aims to improve the capture of subtle mural damage features, thereby enhancing the detection precision. In addition, an Efficient-RepGFPN feature fusion network is implemented. The fusion of low-level features with high-level semantic information enabled the model to extract richer feature information. Additionally, the loss function is optimized through the adoption of Inner-SIoU, which allows the model to focus more on locating damaged areas to enhance precision. The AOT-GAN generative adversarial network is utilized to digitally reconstruct the damaged sections of the grotto murals. The experimental results show that the YOLO Mural model has a detection precision of 72.7%, a recall rate of 60.1%, mAP@0.5 of 63.7%, a harmonic mean of 66%, and a real-time detection speed of 289.32 frames per second. Compared to YOLOv10, the YOLO Mural model exhibited a precision increase of 4.15 percentage points and a recall increase of 10.68 percentage points. In comparison with RT-DETR, YOLO v9, YOLO v8, and YOLO v7 models, the YOLO Mural model has the fewest number of parameters, only 2.65 million. When applied to grotto mural restoration, the AOT-GAN model exhibited a PSNR and SSIM of 34.842 and 0.9832, respectively, providing an important reference for subsequent restoration work and potentially saving a significant amount of labor costs.

Since the underground transportation of coal mainly relies on the mine conveyor belt to complete, the mine conveyor belt with large pieces of coal will affect transportation safety. Therefore, to address the problem of real‐time monitoring of lump coal, the method Ghost‐ECA‐Bi FPN (GEB) YOLOv5 for lump coal in the process of mining conveyor belt transportation is proposed based on a lightweight neural network and multisource information fusion. First, the image preprocessing is performed by adaptive histogram equalization, which reduces the influence of coal dust, dust, and uneven lighting on target monitoring. Second, the redundancy of the convolution process is exploited, and a lightweight neural network GhostNet is introduced to optimize the feature extraction process. In addition, combined with the efficient channel attention mechanism, the 1D convolution enables local cross‐channel information interaction, which can solve the problem of imbalance between model complexity and performance. Finally, the feature information of the three stages is fused using a weighted bidirectional feature pyramid network to enhance the generalization ability of the model. The experimental results show that the improved GEB YOLOv5 algorithm has obvious advantages. In terms of model structure, the number of network layers reduces by 36.97%, and the number of model structure parameters and floating‐point operations reduce by 64.53% and 69.14%, respectively. Moreover, the model volume reduces from 92.7 M to 33.0 M. Regarding the monitoring performance, the precision and recall rates improve by 1.19% and 1.11%, respectively. Furthermore, the real‐time performance improves from 68.34 FPS to 110.70 FPS. It can be seen that the problem of the model performance against the model complexity is effectively solved in this experiment and the real‐time monitoring of lump coal is realized.

Due to its large size of coal and high mining output, lump coal is one of the hidden risks in mining conveyor damage. Typically, lump coal can cause jamming and even damage to the conveyor belt during the coal mining and transportation process. This study proposes a novel real-time detection method for lump coal on a conveyor belt. The space-to-depth Conv (SPD-Conv) module is introduced into the feature extraction network to extract the features of the mine’s low-resolution lump coal. To enhance the feature extraction capability of the model, the normalization-based attention module (NAM) is combined to adjust weight sparsity. After loss function optimization using the Wise-IoU v3 (WIoU v3) module, the SPD-Conv-NAM-WIoU v3 YOLOv8 (SNW YOLO v8) model is proposed. The experimental results show that the SNW YOLOv8 model outperforms the widely used model (YOLOv8) in terms of precision and recall by 15.82% and 11.71%, respectively. Significantly, the real-time detection speed of the SNW YOLOv8 model is increased to 192.93 f s⁻¹. Compared to normal models, the SNW YOLO v8 model overcomes the disadvantages of normal models, such as being overweight, and the parameters of SNW YOLO v8 are reduced to only 6.04 million with a small model volume of 12.3 MB. Meanwhile, the floating point of SNW YOLOv8 is significantly reduced. Consequently, it demonstrates excellent lump coal detection performance, which may open up a new window for coal mining optimization.

North of Shanxi, Datong Yunzhou District is the base for the cultivation of Hemerocallis citrina Baroni, which is the main production and marketing product driving the local economy. Hemerocallis citrina Baroni and other crops’ picking rules are different: the picking cycle is shorter, the frequency is higher, and the picking conditions are harsh. Therefore, in order to reduce the difficulty and workload of picking Hemerocallis citrina Baroni, this paper proposes the GGSC YOLOv5 algorithm, a Hemerocallis citrina Baroni maturity detection method integrating a lightweight neural network and dual attention mechanism, based on a deep learning algorithm. First, Ghost Conv is used to decrease the model complexity and reduce the network layers, number of parameters, and Flops. Subsequently, combining the Ghost Bottleneck micro residual module to reduce the GPU utilization and compress the model size, feature extraction is achieved in a lightweight way. At last, the dual attention mechanism of Squeeze-and-Excitation (SE) and the Convolutional Block Attention Module (CBAM) is introduced to change the tendency of feature extraction and improve detection precision. The experimental results show that the improved GGSC YOLOv5 algorithm reduced the number of parameters and Flops by 63.58% and 68.95%, respectively, and reduced the number of network layers by about 33.12% in terms of model structure. In the case of hardware consumption, GPU utilization is reduced by 44.69%, and the model size was compressed by 63.43%. The detection precision is up to 84.9%, which is an improvement of about 2.55%, and the real-time detection speed increased from 64.16 $FPS$ to 96.96 $FPS$, an improvement of about 51.13%.