This article proposes a hierarchical multiobjective heuristic (HMOH) to optimize printed-circuit board assembly (PCBA) in a single beam-head surface mounter. The beam-head surface mounter is the core ...facility in a high-mix and low-volume PCBA line. However, as a large-scale, complex, and multiobjective combinatorial optimization problem, the PCBA optimization of the beam-head surface mounter is still a challenge. This article provides a framework for optimizing all the interrelated objectives, which has not been achieved in the existing studies. A novel decomposition strategy is applied. This helps to closely model the real-world problem as the head task assignment problem (HTAP) and the pickup-and-place sequencing problem (PAPSP). These two models consider all the factors affecting the assembly time, including the number of pickup-and-place (PAP) cycles, nozzle changes, simultaneous pickups, and the PAP distances. Specifically, HTAP consists of the nozzle assignment and component allocation, while PAPSP comprises place allocation, feeder set assignment, and place sequencing problems. Adhering strictly to the lexicographic method, the HMOH solves these subproblems in a descending order of importance of their involved objectives. Exploiting the expert knowledge, each subproblem is solved by an elaborately designed heuristic. Finally, the proposed HMOH realizes the complete and optimal PCBA decision making in real time. Using industrial PCB datasets, the superiority of HMOH is elucidated through comparison with the built-in optimizer of the widely used Samsung SM482.
Optimizing all the objectives of the printed circuit board assembly (PCBA) optimization in a multifunctional placer remains a formidable challenge till now. This article converts the original PCBA ...optimization problem to a newly defined component allocation problem, which decides the component-type handled by each head per pickup-and-place (PAP) cycle. The component allocation problem is a quadratic 3-D assignment problem (Q3AP) and effectively combines the optimization of all the main objectives. It is possible that one head stays idle, so the assigning 2-D locations are uncertain. We propose the cell division genetic algorithm (CDGA) to solve such a complex Q3AP. The CDGA allocates a component cell as the basic unit. Each of the first-generation component cells contains the mounting points of the same type. A cell chromosome decoding heuristic is designed to determine the next assigning head. By doing so, the problem dimension is reduced, so the conventional GA can be used for searching the optimal component allocation formed by the current-generation cells. When a better allocation can no longer be found by allocating the current cells, the cell division operation is performed to divide each cell into two new cells. The new cells are used in the next round of GA searching, which further optimizes the allocation from two perspectives: better balancing the minimization of nozzle changes and PAP cycles, more flexibly maximizing the simultaneous pickups with the uncertain locations. The CDGA works continuously until the current cells cannot bring any improvement. In simulations and experiments using the industrial samples, the proposed algorithm significantly reduces the PCBA time compared to two recent studies and the built-in optimizer of the widely used multifunctional placer, Hanwha SM482 PLUS, which demonstrates its effectiveness and superiority.
Hopfield neural network (HNN) is a well-studied optimization method, but has not been able to solve the capacitated location routing problems (CLRP). Transporting components from the feeders to the ...placement points by a capacitated head set, the pick-and-place location routing in multi-functional placers is studied as a typical CLRP. The original problem is decomposed into three subproblems deciding the placement points grouping, feeders location and inner-group placement sequencing, respectively. The first two subproblems are optimization problems with multiple tours. With heuristics designed for optimization within a single tour, these two subproblems are transformed into the heuristic sequencing optimization problems that optimize the solving sequence of the tours. The typical HNN for the travelling salesman problem is improved, which results in the heuristic sequencing HNN. The energy function is adapted for reflecting the activation states of the energy matrix about the index-sequence pairs. Through exploration into the objective function value of each index-sequence pair, the energy matrix is calculated by a specially designed method combining with a normalization technique. To achieve optimal performance of the HNN-based methods, a multi-start mechanism is employed. The effectiveness and efficiency of the proposed method are finally elucidated by experiments using practical industrial data.
•The developed system's registration is robust to board size.•Head-mount visualization is appropriate for inspection compared to handheld based.•Multi-stage rendering of images is suitable in the ...vergence-accomodation conflict region.•Superposing position with orientation information provides more efficient guidance.•Contextual displays are more appropriate than static ones during inspection.
Augmented reality (AR) is a key technology anchored towards realizing Industry 4.0 smart manufacturing aims. In manufacturing inspection, AR has previously been employed to support operators in assessing thickness of manufactured parts, ship building, aircraft subassemblies and printed circuit board assemblies (PCBAs) with or without external markers in standalone head-mounted (HMD) or handheld systems. These AR systems often use optical see-through (OST) or video see-through (VST) technologies. Currently, cyber physical integration of processes in an industrial system synergizes production through increased efficiency and improved quality while facilitating customization. An operator manually double-checking/inspecting a product that has previously been automatically inspected can better rely on this existing defect location information once it is contextually and spatially overlaid within their field of view to intuitively and efficiently execute the task. Additional interactive information provided can quickly reorient a user, provide necessary contextual reference, and monitor progress of the task while alerting them of their safety. In this study, adopting an automatic optical inspection (AOI) wirelessly aided HMD-OST AR-based manual inspection system requiring no external markers for contextual registration is a promising direction towards smart and safe PCBA manufacturing in line with Industry 4.0. Animated rectangular bracket of high color contrast is employed to localize the spawned AOI defect point, an arrow overlaid with distance text provides user guidance, a contextual resizable image facilitates user double-checking and a progress bar tracks the inspection progress while monitoring tracking state. We evaluate the developed system's robustness, effectiveness of display technology used, suitability of contextual against static display modes, and display technology influence on various inspection attributes in a user study. Results demonstrate the system to be robust, OST-HMD outperforms handheld VST devices, contextual display mode to be significantly preferred to static mode, and display technology employed has no significant influence on the inspection attributes. Finally, registration precision results demonstrate usability of the system while superposition of distance to orientation information raises the inspection rate of PCBAs.
•A new Bees Algorithm version for PCB assembly time minimisation is presented.•The new Bees Algorithm version is customised for combinatorial optimisation.•The new algorithm outperformed three ...state-of-the-art methods on the TSP benchmark.•The new algorithm outperformed the state-of-the-art on the PCB assembly task.
This paper presents a novel version of the Bees Algorithm customised to solve combinatorial optimisation problems. This version was created to minimise assembly time in the manufacturing of printed circuit boards using a machine of the moving-board-with-time-delay type, and optimising the feeder arrangement and machine component placement sequence. The local search procedure of the standard Bees Algorithm was modified to include five new operators for combinatorial optimisation. The customised Bees Algorithm was first tested on the related travelling salesman problem, where it excelled in terms of performance and efficiency compared to three state-of-the-art optimisation methods. It was then applied to a well-known moving-board-with-time-delay benchmark problem, where it performed favourably in comparison to the state-of-the-art in the literature, achieving fast and consistent solutions.
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
The joint task of allocating several PCB assembly jobs to a set of production lines, load balancing of the line machines and job scheduling is considered. The production facility includes a number of ...assembly lines of different kinds, the PCB jobs are of different types and they should be allocated to suitable (i.e. feasible) lines. Scheduling of the production should respect the predefined release and due dates, and the objective is to minimise the sum of job tardy times. The scheduling is of the rolling-horizon-type where at the beginning of each planning period new jobs are inserted in the current non-preemptive production programme of unfinished jobs from the past planning periods. A mathematical formulation and a two-phase heuristic (including initial job-to-line allocation and schedule improving steps) are given for the problem. Experimental tests with jobs from practice were convincing.
The printed circuit board assembly (PCBA) tends to become a bottleneck in an assembly line of electronic products. For improvement, many assembly firms have introduced chip shooter machines. This, ...however, in turn, raises the issue of how to best utilize these machines. To deal with the PCBA problems, swarm intelligence (SI)-based metaheuristics have been increasingly popular due to the use of guided search that can better drive solutions toward optimality. In this study, we have proposed a novel SI-based metaheuristics, termed improved shuffled frog-leaping algorithm (I-SFLA2), to deal with the feeder assignment problem (FAP), and component sequencing problem (CSP) simultaneously for a chip shooter machine. With novel features such as self-adaptive jump, push jump, direct-jump prevention, and self-adaptive variant, the I-SFLA2 can develop smart jumps for frogs to approach and search around elites quickly while avoiding being trapped in local optima. The I-SFLA2 includes the strategy of transitioning from exploration to exploitation by decreasing the number of memeplexes iteratively. Our small-sized experiments showed the I-SFLA2 had a high hit rate to the optimal solution whereas big-sized experiments showed the I-SFLA2 outperformed the basic SFLA (B-SFLA), B-SFLA(2), the improved SFLA (I-SFLA1), as well as the PSO2 proposed in previous studies. The computational times for the I-SFLA2 were found also reasonable for practical usage. Note to Practitioners-Chip shooter machines have been widely used in industry for printed circuit board assembly (PCBA). Approaches such as exact approach, simple heuristics, and metaheuristics have been proposed for PCBA planning. Although with the capability to find the optimal solution, the exact approaches are found to be computationally intractable when used to deal with a problem of a practical size. Although simple heuristics are easy to use, they usually have the difficulty to find the optimal/near-optimal solution. One recent trend is using metaheuristics to solve PCBA problems as they can avoid the computational intractability of exact approaches while improving over simple heuristics to find a better solution. In this study, we propose an improved shuffled frogs leaping algorithm (I-SFLA2), an advanced swarm intelligence (SI)-based metaheuristic to solve the CSP and FAP simultaneously for a chip shooter machine. With new features and smart jumps, the I-SFLA2 is able to find a better solution compared to some previously-proposed methods.
The dismantling of printed circuit board assemblies (PCBAs) and the recovery of their useful materials can lead to serious environmental impacts mainly due to their complicated physical structure and ...the variety of toxic elements contained in their material composition. So far, less attention has been paid to their responsible recycling compared to that of bare printed circuit boards. Combined with other materials recovery process, proper dismantling of PCBAs is beneficial to conserve scarce resources, reuse the components, and eliminate or safely dispose of hazardous materials. In analyzing the generation, resources potential and hazardous risk of scrap PCBAs, technologies used for the dismantling of waste PCBAs have been widely investigated and reviewed from the aspects of both industrial application and laboratory-scale studies. In addition, the feasibility of PCBA dismantling has been discussed, the determinants of which, including the heating conditions and mechanical properties have been identified. Moreover, this paper evaluates the environmental consequences caused by the dismantling of PCBAs.
A deep ensemble convolutional neural network (CNN) model to inspect printed circuit board (PCB) board dual in-line package (DIP) soldering defects with Hybrid-YOLOv2 (YOLOv2 as a foreground detector ...and ResNet-101 as a classifier) and Faster RCNN with ResNet-101 and Feature Pyramid Network (FPN) (FRRF) achieved a detection rate of 97.45% and a false alarm rate (FAR) of 20%-30% in the previous study <xref ref-type="bibr" rid="ref34">34 . However, applying the method to other production lines, environmental variations, such as lighting, orientations of the sample feeds, and mechanical deviations, led to the degradation in detection performance. This article proposes an effective self-adaption method that collects "exception data" like the samples with which the Artificial Intelligent (AI) model made mistakes from the automated optical inspection inference edge to the training server, retraining with exceptions on the server and deploying back to the edge. The proposed defect detection system has been verified with real tests that achieved a detection rate of 99.99% with an FAR 20%-30% and less than 15 s of inspection time on a resolution <inline-formula> <tex-math notation="LaTeX">7296 \times 6000 </tex-math></inline-formula> PCB image. The proposed system has proven capable of shortening inspection and repair time for online operators, where a 33% efficiency boost from the three production lines of the collaborated factory has been reported <xref ref-type="bibr" rid="ref6">6 . The contribution of the proposed retraining mechanism is threefold: 1) because the retraining process directly learns from the exceptions, the model can quickly adapt to the characteristic of each production line, leading to a fast and reliable mass deployment; 2) the proposed retraining mechanism is a necessary self-service for conventional users as it incrementally improves the detection performance without professional guidance or fine-tuning; and 3) the semiautomatic exception data collection method helps to reduce the time-consuming manual labeling during the retraining process.