Combing Automated Advanced Process Control with Feedback to Revolutionize the Printed Circuit Board Assembly Process

Surface-mounted technology (SMT) welding defect detection plays a key role in the printed circuit board assembly (PCBA) production process, which affects the use of electronic products and cost. Previous works tend to realize defect detection with only defect samples and they assume that there are sufficient defect samples. However, defect samples are usually difficult to collect in real-life scenarios while enough template samples can be easily obtained. In addition, most existing works carry out defect detection based on benchmarks with simple backgrounds of PCBA, which is not suitable for PCBA with complex structures in modern electronic product manufacturing. To address the above issues, we propose a template-guidance Siamese network based on YOLO for SMT welding defect detection (TGSYOLO), which is deployed on a real SMT automatic optical inspection (AOI) system. First, the two-stream structure is introduced to extract deep features in defect images and template images, in which template features serve as guidance knowledge. Then, a template fusion Transformer (TFT) is proposed to model global features between detect and template features in the low-level stage, which could acquire long-range correlations to force the network to focus on potential defect regions. Next, to avoid the disappearance of tiny defect features during deep feature fusion, a multiscale attention feature pyramid network (MAFPN) is proposed to directly fuse defect semantic information from low-level features, which retains detailed expressions of defects through skip connection and obtains compact fusion features. Furthermore, we collect limited welding defect samples based on more complex PCBA backgrounds than previous works through a real SMT AOI system. Experiments on the limited dataset show that TGSYOLO could reach 0.985 of mAP@0.5, 0.885 of mAP@0.75, and 0.984 of F1, which is 0.008, 0.054, and 0.025 higher than other SOTA methods. Also, generalization experiments on the public DeepPCB show that TGSYOLO could still reach the best with 0.991 of mAP@0.5 and 0.89 of mAP@0.75, which proves that TGSYOLO has good generalization performance.

This paper examines the process real-world SMT manufacturers are employing to evaluate, select and implement automated optical inspection (AOI) systems for post-placement/pre-reflow inspection during the PCB assembly process. A case study from a prominent European SMT manufacturer describes the process used to evaluate a post-placement/pre-reflow AOI system. Traditional, algorithm-based AOI techniques are briefly described and evaluated, along with their strengths and limitations. A new AOI methodology, called statistical appearance modeling, and a stereo-vision inspection machine are briefly described and compared to traditional AOI systems. Examples of payback and ROI calculations and analysis are highlighted. The system requirements are specified and the actual results of the AOI performance trial are summarized. Other relevant topics include analyzing GR&R, interpreting SPC data, and implementing post/placement/pre-reflow AOI.

A profile identification system for surface mount printed circuit board assembly

Automated Optical Inspection (AOI) is among the most common and effective quality checks employed in production lines. This paper details the design of a Deep Learning solution that was developed for addressing a specific quality control in a Printed Circuit Board Assembly (PCBA) manufacturing process. The developed Deep Neural Network exploits transfer learning and a synthetic data generation process to be trained even if the quantity of the data samples available is low. The overall AOI system was designed to be deployed on low-cost hardware with limited computing capabilities to ease its deployment in industrial settings.

The drive to reduce cost, improve manufacturing processes and meet higher standards for quality are well-known conditions in the electronics industry. OEMs who have chosen to outsource core steps in the process increasingly look to their electronics manufacturing service (EMS) partners to help them meet these challenges. And as EMS companies develop systems and tools designed to address these needs, they are increasingly turning to software systems that measure yield, monitor work-in-process and enable reduced cycle time. This paper will examine one EMS company's experience in developing real-time, Web-based reporting tools for five specific areas: first pass yield, defect data, control charts, historical data retrieval and run rates. This paper will address the techniques and opportunities in implementing a fully integrated system for real-time control and monitoring of quality at various stages in the printed circuit board assembly process. It will explore various methods for collecting and delivering information in support of EMS operations, and will address strategies for applying Web based reporting at various integration points in the electronics manufacturing process. It will also review a technical architecture that has been affective in interfacing to the company ERP system and how Web reporting can provide detailed information concerning key events such as material movement and backflushing.

The objective of this research is to reduce the defects in the manufacturing of Printed Circuit Board (PCB) Assembly according to the case study company's issue. The defects from this process must be repaired by the operator and the repairing process becomes waste. The defects occur from the PCB assembly with the Reflow soldering process. The Six Sigma approach with the Design, Measure, Analyze, Improve, and Control (DMAIC) methodology is applied to define the problem and make the main problem focused. The defects are collected as data from the Daily Process Confirmation Record. The cause-and-effect diagram is used to clarify the possible factors to examine. Furthermore, the Design of the Experiment is applied to analyze the significant factors that cause the defects. The 2k Centre points design is used to test the significance of factors. It is observed that after changing only one factor, i.e. solder paste's preparation, the defects are reduced from 5.80% to 0.00%. Not only are the defects reduced significantly but also the man-hour is reduced too. Furthermore, it is anticipated that the repair time will be reduced from 116 hours to less than 50 hours.

High-reliability electronics demand full-coverage verification of component integrity, yet conventional inspection methods remain limited in scope and unable to detect subtle or undocumented modifications. This work presents a secure hardware assurance framework that leverages visual AI, specifically deep learning applied to high-resolution images from Automated Optical Inspection (AOI) systems, to detect component-level anomalies in printed circuit board assemblies (PCBAs). The modular system integrates object detection, semantic segmentation, and a second-stage anomaly classifier, enabling detection of unauthorized modifications, substitutions, and structural deviations with high precision. Trained using a bootstrapped procedure on multi-domain datasets, including clean production boards and degraded scrap units, the model generalizes across real-world conditions with over 99% detection accuracy and sub-second board-level analysis times. In light of recent hardware-level cyber threats, such as the discovery of rogue communication devices embedded in commercial infrastructure, the need for scalable, image-based verification has never been more critical. This method transforms existing AOI data into an intelligent layer of visual forensics, enabling manufacturers to detect covert hardware changes and enforce trust in deployed electronic systems.

Selecting products that have been qualified by industry standards for use in printed circuit board assembly processes is an accepted best practice. That products which have been qualified, when used in combinations not specifically qualified, may have resultant properties detrimental to assembly function though, is often not adequately understood. Printed circuit boards, solder masks, soldering materials (flux, paste, cored wire, rework flux, paste flux, etc.), adhesives, and inks, when qualified per industry standards, are qualified using very specific test methods which may not adequately mimic the assembly process ultimately used. It is recommended that products used in combination on a printed circuit assembly be qualified in combination to the extent necessary to provide a full understanding of interactions that may occur in the product. IPC J-STD-001 provides good guidance with regards to process validation testing although said testing is limited to the Appendix of the document. J-STD-001D, Appendix C focuses on process validation via Surface Insulation Resistance (SIR) Testing. The limitation to relying solely on SIR testing is its inherent inability to detect possible assembly issues which could result from employed combinations of printed circuit boards, solder masks, soldering materials (flux, paste, cored wire, rework flux, paste flux, etc.), adhesives, and inks. The authors’ intention with this paper is to describe two Designs of Experiments (DOE’s) that were developed for process validation. The first relies on SIR testing as the sole means of qualifying the final assembly. The second explores the use of a broad range of tests chosen to closely represent the end use application. DOE #1 was based upon the SIR testing procedure as per ANSI/J-STD-004, IPC-TM-650 2.6.3.3A the method which is also specified in J-STD-001. The scope of the test method is to determine the degradation of electrical insulation resistance of PCB specimens after exposure to specified materials. Since during the normal manufacturing process, a board assembly is exposed to a number of chemicals, for this DOE the standard SIR test coupons were prepared in such a way as to mimic a “typical” chemical mix that a board assembly may come in contact with during assembly. This involved applying the appropriate chemicals in the typical process order using appropriate process conditions.

Automatic optical inspection (AOI) is a control process for precisely evaluating the completeness and quality of manufactured products with the help of visual information. Automatic optical inspection systems include cameras, light sources, and objects; AOI requires expert operators and time-consuming setup processes. In this study, a novel autonomous industrial robot-guided inspection system was hypothesized and developed to expedite and ease inspection process development. The developed platform is an intuitive and interactive system that does not require a physical object to test or an industrial robot; this allows nonexpert operators to perform object inspection planning by only using scanned data. The proposed system comprises an offline programming (OLP) platform and three-dimensional/two-dimensional (3D/2D) vision module. A robot program generated from the OLP platform is mapped to an industrial manipulator to scan a 3D point-cloud model of an object by using a laser triangulation sensor. After a reconstructed 3D model is aligned with a computer-aided design model on a common coordinate system, the OLP platform allows users to efficiently fine-tune the required inspection positions on the basis of the rendered images. The arranged inspection positions can be directed to an industrial manipulator on a production line to capture real images by using the corresponding 2D camera/lens setup for AOI tasks. This innovative system can be implemented in smart factories, which are easily manageable from multiple locations. Workers can save scanned data when new inspection positions are included based on cloud data. The present system provides a new direction to cloud-based manufacturing industries and maximizes the flexibility and efficiency of the AOI setup process to increase productivity.

Automatic optical inspection (AOI) systems for drones are increasingly becoming core tools for intelligent operation and maintenance of industrial infrastructure. However, their traditional implementation paradigms face fundamental technical bottlenecks in communication security, collaborative control, and harsh environment perception. This paper aims to provide a systematic review and analysis of research progress in enabling nextgeneration UAV AOI systems through emerging quantum technologies, including quantum communication, quantum computing, and quantum sensing. The article first provides a comprehensive overview of the unique potential and current engineering challenges faced by technologies such as quantum key distribution (QKD), quantum approximate optimization algorithms (QAOA), and quantum imaging in addressing the limitations of classical systems. Furthermore, this analysis delved into the cutting-edge concept of a “quantum-classical hybrid architecture” that integrates quantum modules with mature classical systems, analyzing its integration mechanisms and performance gains across three dimensions: communication, computation, and sensing. The review indicates that while this fusion paradigm can theoretically significantly enhance system security, optimize efficiency, and improve robustness, its practical deployment remains constrained by the environmental sensitivity of quantum devices, high costs, and complex system integration. Finally, this paper offers a forward-looking discussion on future development directions in hardware miniaturization, algorithmic collaboration frameworks, and standardization research within this field, providing a framework for researchers to comprehensively understand the opportunities and challenges in this interdisciplinary domain.

Colored contact lenses have gained popularity in recent years. However, their production process is plagued by low efficiency, which is attributed to the complex nature of the lens color patterns. The manufacturing process involves multiple complex steps that can introduce defects or inconsistencies into the contact lenses. Moreover, manual inspection of a considerable number of contact lenses that are produced inefficiently in terms of consistency and quality by humans is prevalent. Alternatively, automatic optical inspection (AOI) systems have been developed to perform quality-control checks on colored contact lenses. However, their accuracy is limited due to the increasing complexity of the lens color patterns. To address these issues, convolutional neural networks have been used to detect and classify defects in colored contact lenses. This study aims to provide a comprehensive guide for AOI systems using artificial intelligence in the colored contact lens manufacturing process, including the benefits and challenges of using these systems. Further, future research directions to achieve a classification accuracy of >95%, which is the human recognition rate, are explored.

Automatic Optical Inspection (AOI) is any method of detecting defects during a Printed Circuit Board (PCB) manufacturing process. Early AOI methods were based on classic image processing algorithms using a reference PCB. The traditional methods require very complex and inflexible preprocessing stages. With recent advances in the field of deep learning, especially Convolutional Neural Networks (CNN), automating various computer vision tasks has been established. Limited research has been carried out in the past on using CNN for AOI. The present systems are inflexible and require a lot of preprocessing steps or a complex illumination system to improve the accuracy. This paper studies the effectiveness of using CNN to detect soldering bridge faults in a PCB assembly. The paper presents a method for designing an optimized CNN architecture to detect soldering faults in a PCBA. The proposed CNN architecture is compared with the state-of-the-art object detection architecture, namely YOLO, with respect to detection accuracy, processing time, and memory requirement. The results of our experiments show that the proposed CNN architecture has a 3.0% better average precision, has 50% less number of parameters and infers in half the time as YOLO. The experimental results prove the effectiveness of using CNN in AOI by using images of a PCB assembly without any reference image, any complex preprocessing stage, or a complex illumination system. Doi: 10.28991/HIJ-2022-03-01-01 Full Text: PDF

An Automatic Optical Inspection (AOI) system for optical inspection of imaging devices used in automotive industry using an inspecting optics of lower spatial resolution than the device under inspection is described. This system is robust and with no moving parts. The cycle time is small. Its main advantage is that it is capable of detecting and quantifying defects in regular patterns, working below the Shannon-Nyquist criterion for optical resolution, using a single low resolution image sensor. It is easily scalable, which is an important advantage in industrial applications, since the same inspecting sensor can be reused for increasingly higher spatial resolutions of the devices to be inspected. The optical inspection is implemented with a notch multi-band Fourier filter, making the procedure especially fitted for regular patterns, like the ones that can be produced in image displays and Head Up Displays (HUDs). The regular patterns are used in production line only, for inspection purposes. For image displays, functional defects are detected at the level of a sub-image display grid element unit. Functional defects are the ones impairing the function of the display, and are preferred in AOI to the direct geometric imaging, since those are the ones directly related with the end-user experience. The shift in emphasis from geometric imaging to functional imaging is critical, since it is this that allows quantitative inspection, below Shannon-Nyquist. For HUDs, the functional detect detection addresses defects resulting from the combined effect of the image display and the image forming optics.

Automated optical inspection (AOI) has been widely used in industrial Quality Assurance (QA) procedures. Multi-task inspection in high-speed AOI systems is becoming a significant problem in the design. In this paper, the design of an AOI system for E-shaped magnetic core elements is briefly described and several novel algorithms are proposed to realize defects detection by this system. First, this paper proposes a robust k-tSL-center clustering method to classify the interfaces of the element into normal and damaged areas. Second, a modified Active Shape Model (ASM) method is adopted to perform shape distortion detection in real-time. Performance evaluations are carried out on an E-shaped Magnetic Core Image Database, in which all images are captured by the designed AOI system. Experimental results show that the proposed methods are more efficient, robust and accurate than state-of-the-art methods in this application.

In computer vision, shape recovery is a classic problem. The shape from shading problem aims to reconstruct a 3D shape from one or more intensity images. Inspection of solder joint is a critical step in the assembly of PCB (Printed Circuit Board). So, non-destructive inspection technology has been widely used in a production of PCB. In automated optical inspection (AOI) field 2D inspection appears to be more prevalent industry-wide than 3D inspection, but the limit of 2D inspection in some defects is apparent. This paper deals with the reconstruction of the solder joint's surface in PCB based on shape form shading technology. The advantages of SFS method are analyzed firstly. Then, theory and survey of SFS method is introduced. To overcome the disadvantages existing in conventional SFS methods, some improved algorithms are provided and proved by experiments. Experimental results reveal that the proposed method shows practical value in solder joint inspection.