Review of Methods for PCB Panel Depanelization and Methods for Correct Assembly of Electronic Components on PCB Panels

Abstract

Currently, processes related to PCBs (printed circle board), such as depanelization and checking the correct functioning of the boards, are carried out in separate devices. The purpose of this article is to review the literature and analyze trends related to these aspects of PCB panel manufacturing. The purpose of this analysis is to indicate the currently used depanelization methods and methods for checking the correctness of the assembly of electronic circuits on PCB panels. The publications were found in such knowledge bases as Scopus, IEEE Xplore or Emerald insight. In the following article, a systematic literature analysis along with a mapping study is used. This publication provides a review of selected scientific papers found in the above-mentioned databases. Based on these analyses, insights related to future work on both aspects of PCBs were presented. These insights are part of the development of new integrated devices for depanelization and verification of PCBs.

1. Introduction

The modern world is based on electronic devices. They can be found everywhere, from home electronics and household appliances to military and astronomical equipment. All of these devices have one common denominator, namely PCBs. A printed circuit board (PCB) is a combination of a non-conductive substrate (materials such as composite epoxy, fiberglass or another composite material are used for it) and paths connecting electronic components (here copper is used as a conductive material). In PCB analysis, we distinguish between different categories using the criteria of the number of layers and the stiffness of the board [1].

Table 1. Division of PCBs [1].

Division of PCBs
Due to the number of layers			Due to the stiffness of the insert
Single-layer	Double-layer	Multi-layer	Rigid	Flexible	Rigid-Flex

shows a more detailed division of PCBs, while Figure 1 shows the cross-section of PCBs in different layers.

A PCB panel is a PCB board that is connected in the form of an array. PCBs are used to increase the efficiency of the assembly process of electronic components. In addition, panelization is used to standardize the dimensions to the standard of the machines (especially when the component boards have an irregular shape) [2]. PCBs are connected in panels in the most common ways. One of them is the so-called V-scoring method. This connection involves the creation of two parallel and straight V-shaped grooves that separate the individual PCBs. In these grooves, about a third of the material is removed from each side, leaving a thin piece of material. This connection is strong enough to hold the panel together and at the same time easy to cut off. This connection is used only for rectangular or square tiles. The second type of connection is the so-called tab-routing method. It is created during the milling of the material around the insert. However, a fragment of the materials that connect the tiles to each other is left behind. Additional mouse-bites are made in these joints for easier disassembly. This joining is usually used for tiles with a non-standard shape and those that have curves in their outline. Both connections are shown in Figure 2.

The depanelization process is nothing more than separating a single PCB from the entire panel. Often performed by hand with tools such as side cuts or pliers, manual depanelization is unfortunately a risky process and can cause all kinds of damage to the PCB, from broken components to broken circuit paths. In addition, manual depanelization can leave imperfections around the perimeter, which can be a hindrance when mounting to the target enclosure. Such an imperfection is shown in Figure 3. The aim of this article will be to define the method of depanelization and to classify the knowledge on this topic [3,4].

Diagnostics of panels and PCBs is an important process from the point of view of the entire life cycle of an electronic product. This process should not be negligible during the production phase of the device. Here are some reasons why you should perform PCB diagnostics:

• Cost reduction;
• Maintaining industry standards;
• Product quality assurance.

An important concept in PCB diagnostics is the so-called In Circuit Test (ICT for short). These tests check the components mounted on the PCB through the use of test probes. The probes are connected to a system whose task is to measure currents and voltages, which allows for verification of the assembly of elements. Due to the small size of some components, test probes are placed in special places in the PCB called test pads. Figure 4 shows a cluster of such test pads.

The following literature review is intended to serve as a theoretical introduction to the research and development work on the design of tools and methods for panel diagnostics as well as for PCB depanelization.

The following article is divided into several chapters. In addition to the introduction, the article presents a chapter on the literature review methodology (Section 2). Then, in Section 3, the results of the literature review are described, broken down into appropriate categories (more on this division is provided in Section 2). The next chapter summarizes the results of the literature review and predict future analyses (Section 4). The last chapter collects conclusions from the entire publication.

2. Review Methodology

To discuss the articles in the following review paper, an appropriate strategy for searching knowledge bases is used, which is a key element of analysis in the scientific field. This strategy was created based on several other review works. Three of them describe a literature review for the topic related to digital twin. Each work focuses on a different aspect of this technology. The work of Lo’s team [5] describes a literature review regarding product design and its development for digital twin. Another work [6] is also based on a literature review for digital twin, but in this case, it focuses on concepts and applications in the industry. The last work from this trio is also a literature review; however, it focuses on the product-service system in relation to digital twin [7]. The authors of this work also leaned towards predicting trends of the discussed technology at the end of their study. The last work, which served as an inspiration for the review methodology, describes multidisciplinary design optimization in the context of aircraft [8]. Here, the authors also allowed themselves to predict possible trends. The above works were an inspiration for the following review methodology and for the entire article. It utilized knowledge about keywords, possible places where one can review publications and articles. Additionally, it was also inspired by the possibility of presenting data and the structure of the articles themselves. During the search, it was decided to use the well-known and widely available databases Scopus, IEEE Xplore, Google Schoolar or Emerald Insight. These knowledge bases have been used due to their ease of use and universal access for the user. In addition to standard knowledge bases, ChatGPT 3.5 was also used for the search, which was used to formulate some definitions. The types of scientific papers searched in the above-mentioned search systems were limited to articles in peer-reviewed journals and conference articles. The article has been prepared in order to look at the current trends in PCB panel testing as well as cutting them. An additional intention was to indicate a probable prediction as to the development of this branch of electronic device manufacturing.

The first iteration of selecting keywords for searching for scientific papers was based on the authors’ experiences. Papers initially found only partially covered the discussed topic. It was then decided to expand the keyword base based on previously found publications. After their analysis and joint discussion, a few more keywords were added that could contribute to increasing the number of publications. The most important keywords were tried in the search engine in a changed grammatical form. The results of such a procedure, for example for the pair of words “depanelization” and “depaneling”, were different. However, the found papers were not related to the researched issue. A summary of the selected review parameters is shown in

Table 2. Parameters selected during the literature review.

Exploration Parameters	Title, Abstract and Keywords
Keywords	“automation”, “pcb panels”, “depanelization”, “depaneling”, “cutting pcb panels”, “electronic circuits”, “production processes”, “automated separation process”, “test probes”, “diagnostics of pcb plates”, “pcb panel connections”, “spring contact probe”, “ICT”, “optical inspection”, “flying probe”, “saw cut”, “punching”
Time span	from January 2000 to January 2024
Source type	Journal articles, conference proceedings

.

The literature review consisted of several important stages. The first stage was to search the above-mentioned knowledge bases using the already mentioned keywords and phrases. Subsequently, a preliminary selection of articles was made by reviewing the abstracts of these publications. This step made it possible to exclude items that did not fit the discussed topic and to limit the number of references. The next stage was to read the full texts of the selected articles and define their subject matter. Some of the publications were so interesting that during their analysis they were used to refer to the works included in the references. At the same time, it turned out that the works from references broadened the issue, which resulted in their being added to the database. These stages allowed for the clarification of the thematic categories needed for further analysis. Seven main thematic groups have been distinguished:

• Diagnostics of electrical boards by means of automatic vision;
• Diagnostics of electrical boards using the so-called “bed of nails”;
• Diagnostics of electrical boards using the so-called “flying probe”;
• Other forms of diagnostics;
• Cutting panels and tiles by means of laser processing;
• Cutting panels and tiles by milling or drilling;
• Other forms of PCB depanelization;
• Other publications (related to PCBs and the manufacturing process, not necessarily to the other categories).

The last stage of the review of scientific publications was to conduct an analysis based on the characteristics of the literature such as the time of publication, content of the literature, and its life cycle; in the following subsections, an analysis of the individual categories is presented. In addition, during the search for publications, a database of all publications that referred to depanelization methods or PCB diagnostic methods was created. This database was created in Microsoft Excel (version 2402), which made it easier to filter sources and make charts.

3. Results of the Literature Review

During the first phase of the search, 291 publications were obtained, focusing mainly on scientific articles and conference proceedings. This number included articles describing processes related not only to PCB panel diagnostics or depanelization, but also those that discussed PCB manufacturing and recycling processes. For this reason, an additional selection of materials has been carried out using the keywords described in the second chapter of this article. The final number of publications is 42. See Figure 5 below, in which the quantitative values of each of the categories presented in the previous chapter are shown.

3.1. Diagnostics of Electrical Boards Using Automatic Vision

Nowadays, a lot of attention is paid to the visual inspection of manufactured parts. This form of product verification has become a permanent fixture in the production features of companies. Visual inspection or automatic vision can be found in companies from the automotive industry, as well as in companies dealing with the food industry. The electronics industry also uses vision inspection. One example of such an implementation is the publication of Mohammed A. Alghassa [9]. This is one of the more recent papers discussing the use of automatic vision to verify the correct assembly of printed circuit boards. In it, he describes the use of convolutional neural networks to detection damage on PCBs. The main idea of this method is to introduce a large amount of data into a deep neural network, in which pairs of photos of damaged and undamaged PCBs are introduced. Convolutional neural networks were used to segregate the test data. The author of the publication compared the results of his data with the most modern and state-of-the-art technique and proved that his solution is superior to the previous ones.

Another example of the use of a neural network to inspect printed circuit boards and their circuits is the work of Mariusz Sikora and Michał Grochowsk [10]. The solution of the authors of the publication was created in the MATLAB environment. In addition, they used a simple digital camera to take pictures of the test subject. This publication describes the image processing process according to the following steps:

• Taking a photo with a camera;
• Pre-binarization of the image;
• Cropping the photo to the test plate;
• Binarization of the blue color;
• Using a median filter;
• Superimposition of a reduction grid on the examined image.

In the described case, information in the form of a single pixel is provided to the neural network at the input. However, this solution increases the amount of data to be processed by the algorithm, which in turn increases the time of image analysis. For this reason, the authors of the article divided the photo into larger fragments called an eyelet. They carried out research aimed at selecting the optimal mesh size in terms of the amount of data and the quality of the solutions obtained. The authors of the publication also paid attention to the quality of light used for the work of the research system.

A different approach to automatic vision was proposed by Mengke Li’s team [11]. In their paper, the team describes the possibility of using infrared thermal imaging to detect defective components of printed circuit boards. The entire method was based on infrared thermal imaging, which is part of a multi-sensor detection method. The authors of the publication observed that the various aberrations formed on the PCBs differ significantly in the thermal radiation between the background. Taking advantage of this fact, they designed a network that was trained on a set of images covering various defects of printed circuit boards, such as the following:

• Hole;
• Mouse bite;
• Short;
• Open;
• Spur;
• Spurious copper;
• Exposed copper.

The above-described solution has been experimentally tested, and its results have been compared with other image acquisition models. The authors’ solution turned out to be better than other methods, which they proved by calculating such parameters as the following:

• Average precision;
• Mean average precision;
• Intersection over union.

3.2. Diagnostics of Electrical Boards Using the So-Called “Bed of Nails”

Automatic vision is one of the most popular methods of panel diagnostics. However, we are not able to use this method to diagnose electronic circuits in terms of their functionality. It may happen that the components mounted on the PCB simply do not comply with the bill of materials, especially if the electronic device has several dozen smaller elements. That is why it is a good idea to carry out tile diagnostics with a bed of nails.

Bed of nails tests are custom tests of electronic components to detect manufacturing defects. These tests are performed using an additional electronic circuit. Most often, voltage states on individual components or specific parameters dependent on the element are checked in this way. These measurements are made using appropriately designated locations in the perimeter of the device (so-called test pads). Test probes made of the right material and with the right tip are applied to these areas. They pass the test signal through the aforementioned pads and check the functional aspect of the component [12,13].

During the review study, several interesting articles related to bed of nails tests were found.

One of them describes the process of designing a tool for programming small printed circuit boards. The authors of the text have designed a tool that allows you to program an electronic circuit using bed of nails. They propose a system consisting of several LEDs, a board with MSP430 microcontroller and, of course, test probes. To ensure that the needles are constantly pressed against the surface of the test fields, the authors of the publication designed a simple pressure mechanism. The solution was also aimed at inventing a system that would not require it being connected to a compute [14].

Unfortunately, the previous work did not touch on a rather important issue when testing PCBs. We are talking about the accuracy of the positioning of the needles in relation to the test pads. This aspect is discussed in the Reliability Study for Test Lands Targeting during Electrical Testing [15]. According to the authors, there are several reasons why test needles do not digest perfectly in the center of the test pads. One of them is the deformation of printed circuit boards due to uneven distribution of the pins. The authors also point out that there is a stricter approach to controlling the size of printed circuit boards as well as the height of components.

Another important problem when using bed of nails is the stress on the PCBs caused by the uneven distribution of the test probes. A tool that has a poorly designed such distribution can cause damage to SMT components by cracking the solder joints. These causes are mentioned by the authors in the article Easing PCB stress from bed-of-nails. The authors suggest that the tensometric sensor for stress analysis should be localized not only for important components such as BGAs and CPGAs, but also for low-cost components such as condensers. They noted that a slight change in the position of the sensor can lead to significantly different measurement results. The authors propose to perform earlier stress tests in order to make corrections early enough [16].

3.3. Diagnostics of Electrical Boards Using the So-Called “Flying Probe”

The PCB testing method with the “flying probe” has been known since 1987. The method is based on the use of test probes, which are placed on movable arms (high positioning accuracy). These arms move at very high speed between test points (test pads) and perform simple measurements between individual points. Current solutions use up to 6 test arms, which gives great flexibility during the diagnostic process. The “flying probe” method is used for devices with a small volume (prototype or low-volume production).

Although this method is almost 40 years old, methods are still being developed to help improve this method. In an interesting article by Russell, the accuracy with which the “flying probe” machine positions itself on the test pads is examined. In addition, the author discusses in his paper the damage to the test probes themselves and checks the positioning accuracy with the probes being warped [17].

It is not just the positioning of probes that is being studied by scientists. One of the most recent papers discusses a method for determining the paths of probes. This is to minimize collision with components on the board or PCB panel. The author of the work, Li, describes all the principles and the course of the discussed method [18]. Another example of optimizing the movement of test probes is the article presented by Hiratsuk [19]. Although the article presents an optimization method only for “flying probes” with two moving probes, the authors were able to propose an algorithm that minimizes the movement of probes from one test point to the next.

Some works also focus on improving the performance of the test itself, for example, two works by Boneria. The first paper discusses the possibilities of optimizing test generation and reducing the time needed for measurement [20]. In the second paper, the authors propose improvements to the mobile probes in the SPEA 4080 machine. The aim of the study is to shorten the test time along with the test generation [21].

Part of the work focuses on the possibilities of extending the functionality of the “flying probe” test machines. Usually, it is the use of automatic optical vision, as Deaves did in his work [22]. Radev and his team, using the vision inspection module, describe the possibilities with which they were able to extend the “flying probe” method [23].

3.4. Other Forms of PCB Diagnostics

The forms of PCB diagnostics presented in the previous two subsections are not the only ones that are used in production processes. Other forms of diagnostics have been identified as those that cannot be assigned to the above categories. Below, we will present a few publications that will describe a different way to assess the correct operation of PCBs.

Throughout the production cycle, printed circuit boards are transported by a series of machines that perform specific manufacturing operations, such as applying solder paste or assembling components. During these processes, the PCB is exposed to too much stress, which can cause damage. This can result in cracked SMT-assembled components or solder layers separating. However, these stresses can be investigated and appropriate intervals can be determined to determine whether the inserts are suitable for further processing. The authors of the publication PCB Tests during Assembly and Splitting [24] used the JEDEC methodology [25] to perform tests to check the stress acting on the PCB. They postulated that attention should be paid to the way the sensors are mounted in such a way that the strain gauge cables do not interfere with the production process. The methodology described in the article also required that the individual processes do not destroy the sensors, which is why, for example, the elements that are threaded through the laminate were applied manually. The method created by the authors of the text, after some changes, could be used to test PCB panels as well as individual boards in order to detect irregularities.

Another method of PCB diagnostics is also the verification of the integrity of the board layers and its metallized holes. The method by which these elements can be checked is called interconnect stress testing (IST). The IST method uses specially designed connections on the PCB through which direct current flows. The flow of current causes changes in temperature and resistance, which is observed and monitored. The IST system works by cycling heating and cooling the connections. First, the plates are heated for three minutes and then cooled by an air stream for another 2 min to ambient temperature. During this cycle, joints are examined to identify damage. Interconnect stress testing allows you to test various levels of connections independently, which helps to identify mechanical, individual or interactive failures. This system can detect possible changes in metallization, which can then be translated into an assessment of the integrity of the joints over multiple cycles [26].

In the literature, in addition to methods testing the resistance of PCBs to mechanical or thermal stresses, you can find articles in which the functionality of the boards itself is tested. In this case, automated test equipment is most often used. This refers to automatic tests whose task is to perform diagnostics of the functioning of printed circuit boards. ATE systems are diverse and depend on the application of the board itself. For example, it could be a system to check the performance of the refrigerator’s electronics [27] The LabVIEW program was used to design a graphical interface showing the individual steps of the test. A suitable test board with components such as the ATmega 16 processor and the 16x2 LCD display was also designed for this system.

Yet another, but equally interesting, method of diagnostics is the detection of damage using acoustics. The authors of the publication A System for Detecting Failed Electronics Using Acoustics [28] used 2 MHz ultrasound to verify the damaged components mounted on the PCB. The damaged components were 1 kΩ resistors and various types of integrated circuits. Overvoltage faults were caused in all components. The results of the study show that the method proposed by the authors has a high diagnostic efficiency which, depending on the examined element, ranges from 50 to 85 percent. These are promising results, especially when we mention the low cost of the system.

3.5. Cutting Panels and Tiles by Milling or Drilling

During the literature review, several solutions were found covering the issue of cutting PCBs by means of mechanical processing.

Parts of the publication referred, for example, to the analysis of cutters and the classification of their failures [29]. On the other hand, the quality aspect of drilled holes was touched upon in the work of Xiaohu Zheng’s team and his team. They performed a variance analysis for hole quality parameters. The main conclusion of their work was to indicate the causes of imperfections in the drilled holes, and nine drill bits with different types and point angles were used for the study. The experiment conducted by Xiaohu Zheng’s team showed that the operating parameters of the spindle and the drills themselves are especially important and should be taken into account when performing machining on PCBs [30]. One of the most interesting articles was the one discussing the mathematical modeling of the process of cutting electronic components on PCBs. This modeling was to be used in the design of a machine that would perform such removal of components and contribute significantly to the facilitation of PCB recycling [31].

Several of the articles found were about designing and building PCB panel depanelization tools. The article that deserves the most attention describes the method of designing using a morphological matrix. It shows different solutions and each of the machine’s functionalities. The authors of the publication also performed a quality function deployment analysis in order to obtain information on the requirements for these types of tools [32]. Another interesting article is Design and manufacturing of a PCB cutting machine. It mainly discusses calculations related to the mechanical design and selection of components for the machine to cut PCB panels with a circular saw. The calculations in the publication are reliable and supported by appropriate standards [33]. An interesting initiative in the design of printed circuit board milling machines came from the team of Ms. Kubra Fathima. They designed a CNC milling machine with low-cost components and high availability on the market. In their paper, they described the specification of such a machine, as well as a block diagram of how the tool works [34]. The articles described in this category did not refer entirely to depanelization, but the solutions contained in them can be used for this process.

3.6. Cutting Panels and Tiles Using Laser Processing

Not only typically mechanical solutions such as milling or drilling are used for the processing of PCBs, but also laser cutting forms. A laser can be called a device that uses the emission of electromagnetic radiation with specific properties. It works on the principle of excitation and emission of photons, causing their concurrent radiation in one phase and wavelength [35].

The first publication relates directly to the depanelization of PCB panels. In this study, research was carried out to check the influence of various parameters of the laser system on how they affect the change in the temperature of the substrate. In their research, the authors designed a mathematical model of the experiment and compared it with the actual results. In addition, the temperature of the substrate was checked depending on the laser cutting distance. During the research, an analysis of the elemental composition of the substrate after cutting was also performed. According to the authors of the publication, laser cutting can improve many aspects related to the properties of PCBs, one of them being the preservation of rigidity, as they are not subjected to any external forces causing stress, as in the case of milling or scavenging. Another key issue is the fact that the temperature of the substrate during the entire process was not higher than that during the soldering process. This demonstrates the marginal effect of temperature on the printed circuit boards [36].

Another interesting work was shared by a group of researchers from the Southern Taiwan University of Science and Technology University. They conducted research related to the selection of parameters of a green laser with a wavelength of 532 nm and diode-pumped solid-state (DPSS) technology. In addition to standard PCBs, the researchers also used so-called flexible PCBs for their research. For the analysis, the authors used such laser parameters as the following:

• Power (W);
• Speed (mm/s);
• Pulse repetition rate;
• The number of cuts.

The selection of the best parameters was based on testing the edges of the cut details. The authors chose one set of parameters for each of the materials, which, in their opinion, damaged the edges of the tiles [37].

In the paper by Dongsig Shin’s team, the focus was only on FPCBs; however, an ultraviolet spectrum laser was used to cut it. As in the previous work, the experiment focused on selecting the best parameters such as power, speed or cutting material. In their study, the authors not only performed thermal analysis and material analysis, but also checked the rate of ablation. They postulate that the collected results will be useful in future work on modern electronic equipment [38].

3.7. Other Forms of PCB Depanelization

During the literature review, several papers describing methods of depanelization other than the above-mentioned milling and laser cutting were found. In this chapter, the articles that deserve the most attention will be described.

A remarkably interesting solution is presented in the article Design of Printed Circuit Board Production using Water Jet Technology. This work is based on theoretical considerations within the design of a machine that could be used to manufacture PCB panels and depanelize them. The technology mentioned by the researchers is based on the use of water with extremely high pressure (up to 600 MPa) as a cutting tool. This method uses special nozzles through which a highly kinetic liquid flows. These nozzles are most often mounted on Cartesian robot-type machines, which use CNC control. As the authors of the paper mention, water jet technology is clean for the environment. They also add that high-pressure water without an abrasive is mainly suitable for soft materials or those with low thickness. The authors of the publication conclude that the creation of such a machine could contribute to reducing the cost of PCB production. Additionally, they claim that this method reduced the amount of chemicals used during the standard cutting process [39].

In another article, the authors noted the need to fully automate the depanelization process, which uses low-thickness circular saws. They found that the solutions used in the industry are inefficient and dangerous. Researchers attempted to design a machine that could handle three aspects at the same time:

• Unpacking the panels (leading them to the saw);
• Cutting panels with a circular saw (transport of panels from the place of loading to the place of cutting);
• Laying and packing of cut tiles with transport.

In their presentation, the researchers also decided to show the process of selecting the parameters of the cutting-off wheel, as well as the solution for unloading and loading. They based the control of their machine on Siemens PLCs. It is worth noting that the machine proposed by the authors of the publication is not very flexible. They have selected extremely strict requirements that can only be met for a section of PCB panels used in production processes [40].

The next work focuses on the analysis of the operation of the hole punching machine in the FPCB. Machines of this type are ones that can again be used to perform the depanelization process. In the publication, the authors made a mathematical model of the cutting machine. In addition, they undertook an analysis and proposed improvements related to the design of the machine. One of them was the choice of the servo motor motion curve in the form of the letter S. Unfortunately, the authors of the publication noted that such a curve profile with a fixed period is not suitable for use in the cutting mechanism. To confirm this defect, the researchers drilled several holes with varying periods of constant velocity. For the experiment, they used an HP 35670A dynamic signal analyzer and a 353B16 PCB accelerometer. The accelerometer was mounted on the cutting head. The holes they made were of mediocre quality, which confirmed their assumption about the wrong choice of curve. This work shows that the use of punches for the depanelization of PCB panels can be a noticeably big challenge when the esthetic edge of the cut is a key factor [41].

The above publication may be supplemented by an analysis performed by the team of Piotr Czyżewski. In their work, they simulate the wear of the punching machine’s punching machine. The study was based on the use of a modified die to which a piezoelectric transducer (204C) was mounted. The researchers’ analysis showed that punch wear (increase in the radius of rounding at the edges) largely defines the final dimension of the cut part. This aspect is particularly important if this technology is to be used for PCB depanelization [42].

4. Results Analysis

In this part of the article, the analysis of the results of the literature review will be discussed. The aim of this literature review was to present the methods of depanelization and diagnostics of PCBs and panels. This article showed that there are many scientific papers available on this subject, but only a small number of them can be used to design methods and machines for depanelization and diagnostics of PCB panels. The division of the articles into appropriate thematic categories allowed the systematization of the state of the knowledge and technology in the studied area. Each of the categories gives you the opportunity to approach the topic in an interesting and unconventional way. It has been noticed that in a given category, articles complement each other’s subject matter, filling in thematic gaps. One paper describes a problem related to the quality of PCBs after laser punching, while another describes the selection of laser operating parameters. There were also many similarities in the articles due to the approach to solving a given problem. For example, all of the automated vision work on panels and PCBs used neural networks to verify imperfections, but each work used a different preview tool. It is worth noting that although nowadays there is a lot of emphasis on ecology, the most popular method of depanelization is still the use of milling processes. Although the publication analysis covered a period of more than 24 years, it can be noted that most of the selected works were written in the second decade of this century. This condition is shown in Figure 6. The largest number of papers were published between 2016 and 2020, representing over 35% of the total number of publications. The graph also shows that after 2005, the number of publications related to PCB depanelization and diagnostics increased rapidly. This may indicate an increased need for research into these methods for the market and industry. The number of publications increased between 2016 and 2020. It should also be noted that by 2030, the number of publications is likely to increase and surpass the total number of publications in the second decade of the twenty-first century.

5. Conclusions

The use of the review methodology described in the second chapter allowed for the creation of a comprehensive literature review. This review may help in the future to design a tool for depanelization and diagnostics of panels and PCBs. The methodology described here can also serve as an example and model for a similar review for less experienced scientists, and not only in the field of electronics. An interesting fact is the use of ChatGPT to look for definitions of the problems studied. Although version 3.5 with free access was used for this task, the search results were very satisfying. In the future, it is worth taking a closer look at this issue and conducting a thorough comparison of the literature review between standard search methods and the use of language models. The researched topic of PCB depanelization and diagnostics is extensive and touches on many aspects and branches of science. This article can be the basis for interdisciplinary discussions in the field of electronics, computer science and industrial automation. As mentioned at the beginning, the article will be used to develop a theoretical introduction to the design of a tool that is to perform two functions: panel diagnostics and their depanelization. The literature review developed here has systematized knowledge on methods for PCB separation and testing. A few solutions have been noticed that will be perfect for the design of the future tool. It is likely that the project will use depanelization methods based on laser processing, while the diagnostics of the panels may be based on methods using bed of nails.