PCB industry urgently needs AI and machine learning
Nowadays, PCB has developed to a completely new stage. New technologies such as high-density interconnect (HDI) PCB and IC substrate (ICS) are introduced, making the entire production process from manual to fully automated. With the further development of manufacturing technology, the process becomes more and more complicated, defect inspection becomes more important and more difficult, and these fatal defects may lead to the scrapping of the entire PCB board. For PCB manufacturing, opportunities are emerging for leveraging artificial intelligence (AI) and optimizing production processes and ultimately optimizing the entire PCB manufacturing process.
PCB manufacturing often relies on experts who have accumulated knowledge for many years. These experts know and understand every step of the manufacturing process, and they know how to use their knowledge to optimize production and increase yield. Human limitations (including misoperations and fatigue) hinder efficiency growth. Operator error or misidentification of PCB defects ("false alarms") can affect yields due to overprocessing and even damage the PCB itself. By integrating AI into the manufacturing process, machines can add value by taking over certain "learned" tasks, while human experts continue to take on more complex tasks that require thinking and interaction while optimizing and "training" Artificial intelligence system. The combination of human and artificial intelligence improves overall efficiency and operations, and is the biggest opportunity for AI systems.
Artificial Intelligence and Industry 4.0
The ultimate trend of PCB development is to have factories that fully integrate the Industry 4.0 system, which uses AI technology at the global and manufacturing system levels. The "global" level includes all systems in the plant, not just a single manufacturing system. Industry 4.0 provides an automation and data exchange infrastructure that enables real-time production analysis, two-way communication and data sharing, traceability, and on-demand data analysis. Within any particular factory, AI can use data obtained from various manufacturing systems and machines to improve processes, which are collected through Industry 4.0 mechanisms (e.g. traceability, two-way communication). Factories benefit because AI analyzes a large amount of system-wide data to optimize plant setting parameters and achieve the highest levels of productivity and yield. Artificial intelligence analysis and self-learning are ongoing and are performed through artificial neural networks. Within a few years, it will eliminate manual operator intervention and lead to the establishment of fully automated factories.
This new PCB manufacturing model requires full connectivity of all factory systems and AI as a monitoring and decision-making mechanism. Currently, there are proprietary and technical challenges that limit the full automation of PCB factories, but AI has been added to individual systems as much as possible, such as automated optical inspection (AOI) solutions. The advantages of moving production facilities to a global AI model include more reliable notification of PCB defects-"real defects" and a feedback mechanism that can identify the root cause of the problem and then automatically modify the factory process to eliminate related problem defects.
A subset of AI, including machine learning and deep learning, will move PCB factories towards the goal of full automation. The algorithms used by machine learning enable computers to use data and examples that they have experienced and learned from to improve the performance of tasks without explicitly programming them. In terms of PCB manufacturing, machine learning can increase yields, improve manufacturing operations and processes, and reduce manual operations, while helping to drive more efficient processing of plant assets, inventory, and supply chains.
Deep learning takes AI to a more sophisticated level, which is beneficial at the level of global factory systems. Deep learning is inspired by the ability of human brain neurons, multilayer artificial neural networks to learn, understand, and infer. In a PCB factory, software systems can effectively collect data and learn from complex representations of patterns and contexts. Learning will then form the basis for automated process improvement in PCB manufacturing.
The implementation of machine learning and deep learning provides PCB manufacturers with the ability to transcend human understanding; artificial intelligence systems discover new optimization opportunities by digging deeper in places where people are unwilling to explore. The AI expert system is very efficient. By using more and more complex parameters to monitor factory systems worldwide, it reduces the number of manual experts required and improves efficiency and best practices.
Utilizing Industry 4.0 sensors (sensors that can send data from the device) and systems, throughout the PCB manufacturing process, from simple read and write functions to advanced tracking of process parameters, up to the smallest PCB unit, can be created globally data. Process parameters can include etching, resist development and even chemical material concentration during manufacturing. Use deep learning to analyze these types of data to inform optimized manufacturing methods and parameters, identify patterns, and make informed decisions about changes needed in the process. All of this can be done 24 hours a day, 7 days a week.
At the system level, such as in the AOI process, AI implementation in the PCB manufacturing shop has a considerable impact on productivity and yield. In this case, machine learning greatly reduces human error when detecting PCB defects. Examples of PCB defects include short circuits and open circuits, and even excess copper is fine. Automated inspections can detect small defects that may not be found by manual inspections or missed by human error. This is the natural result of repeated work.
Without using AI, a classic inspection of 100 panels will usually find 20 to 30 defects per panel, about 75% of which are false alarms. Because the policy requires that all defects be checked manually, the review of false alarms wastes valuable production time, increases the processing of PCB, which may cause new damage and may affect further errors made by operators during the review process.
By performing machine learning on the AOI system, such false alarms and repairs can be greatly reduced. Fewer false positives means less processing on the PCB and it also improves efficiency. In addition, AI provides consistent (dynamically improved) defect classification without the limitations inherent to operators, providing more reliable results and reducing verification time. According to Orbotech internal research, AI in AOI systems has been found to reduce false positives by up to 90%. AOI is unique in that the system can collect more data than any other manufacturing solution, which makes it well-suited as the first step in AI implementation. At the same time, the AOI room is the most labor-intensive area of the PCB factory. Therefore, the use of AI in its process will bring the greatest benefits. For PCB manufacturers, all this means that millions of defects can be more accurately identified and classified, potentially increasing yields and reducing costs.