Industry 4.0 promises huge gains in factory data automation efficiency and productivity, thanks in part to parallel advancements in automotive, the internet of things (IoT) and artificial intelligence (AI). Looking to the future, we can envision industrial processes where machines will be responsible for specific “learned” tasks, exploiting closed feedback loops whereby highly detailed production data is monitored and reported in real-time, and detected deviations automatically trigger source parameter adjustments. In this manner, the systems will control and correct themselves, and yields increase substantially.
This process will generate staggeringly large volumes of data to be harnessed and analyzed — a daunting challenge to manage, but utterly necessary for machine learning (ML), where AI systems are designed to absorb and “learn” from every last bit of available information. Although the electronics industry is some way off from achieving this level of automation and intelligence, the evolution toward Industry 4.0 is well underway, as evidenced by recent trends in the PCB manufacturing market.
PCB manufacturers are under intensifying pressure to provide increasingly detailed data on each and every finished PCB unit produced on their shop floor, enabling their customers to quickly identify, track and troubleshoot PCB defects throughout the supply chain with the highest levels of data granularity. To meet this challenge, PCB manufacturers are undertaking process modifications that capture PCB production data and root cause analyses at every stage in the manufacturing process. The process generates vast databases of manufacturing data that ultimately help to deliver full visibility into the manufacturing process, assess functional integrity at each individual layer within a given PCB, and monitor the device’s status as it progresses through the production line.
In addition, PCB manufacturers are seeking methods to use smart manufacturing in new ways to improve the overall process and identify problem areas in factory systems. Advanced process control and visualization applications are key components of this strategy and ultimately will become the enabler for fully automated factories.
Precision tracking at every layer
The benefits of traceability and process control capabilities are numerous and very much aligned with the core tenets of Industry 4.0. Traceability – the ability for manufacturers to track defects all the way to the PCB unit – provides increased process visibility through the gathering of digital data throughout the production line and the channeling of this data directly to the factory’s manufacturing execution system (MES) and IT department. In addition, traceability boosts efficiency through the establishment of a centralized connection point for all manufacturing equipment, and it is cost effective, particularly in the way it leverages analytical tools that enable improved yield and process management.
In practice, PCB traceability is not dissimilar from commercial merchandise and material tracking processes insofar as it’s reliant on barcoding and software. But unlike most commercial goods, PCBs are vastly complex, and the implications of a PCB failure in the field can be particularly severe, depending on the nature of the electronic device it’s housed in, be it a smartphone, defibrillator or unmanned vehicle — and everything in between.
Figure 1. PCB manufacturers need to collect and track data.
It’s not good enough to simply read and track a barcode off the top of a finished PCB because the traceability will be very limited, as will the ability to pinpoint problems that stem from an interior PCB layer. Instead, in the spirit of Industry 4.0, every unit of every layer in every PCB must be individually coded using sophisticated barcode marking and software so that it can be logically linked to the other units in the PCB’s layers.
This extremely detailed approach allows for end-to-end traceability and enables root cause analysis down to each specific PCB manufacturing process within the production line. Among the granular data that can be tracked: machines used and their parameters and actions, pictures of defects, date and time of manufacturing, operator name, lot number, and more.
This tracked data is valuable for historical analysis, but it’s equally valuable – or more so – when it’s processed, analyzed and immediately fed back to the MES before the PCB ever leaves the factory. This data needs to be assessed and acted on in real time if it’s to be used effectively to identify and correct process issues as they happen, thereby limiting the propagation of defects. And as ML and AI technology comes to the fore, this real-time closed feedback loop will prove essential for enabling truly automated learning and decision-making processes.
Deeper understanding of manufacturing processes
There’s a growing need to have full visibility into PCB manufacturing processes, going well beyond the traceability of PCB units. Until now, no automated decisions could be made and operators on the production floor would never be able to look back in time across many manufacturing systems to identify problems and trends accurately. The advent of advanced process control and visualization marks the first steps toward a fully automated factory enabled by AI.
Using advanced process control and visualization, automated and highly detailed production reports will provide manufacturers with a big picture, early in the process – real-time visibility and a defect distribution map across the manufacturing and inspection stages, as well as from the panel level down to individual PCB units. PCB manufacturers can quickly and accurately identify defect trends throughout the design and manufacturing process using analytics collected through production monitoring.
Based on actual production data that provides actionable insights, PCB manufacturers can then carry out root-cause analysis and close the feedback loop with designers. This improves overall production floor management, resulting in efficient, fast and more informed decisions that improve the production process. Benefits can include lower maintenance costs, reduced downtime, and higher productivity and effectiveness.
Gathering data, gathering speed
As painstaking an effort as traceability requires, it is becoming the new norm for embedded PCB manufacturers as they compete for advantage in lucrative electronic device markets with little to no tolerance for defective products. Device integrity and traceability wield an impact far beyond product reliability, however. Indeed, a PCB manufacturer’s ability to demonstrate and maintain high levels of product quality and yield will be an increasingly important factor for customers vetting a highly competitive field of PCB suppliers. Electronic device OEMs are applying more and more scrutiny to suppliers’ traceability processes and the subsequent impact on yield calculations, which in turn informs customers’ assessments of suppliers’ production scalability and cost structures. QR codes (Figure 2) let manufacturers track each board.
At a higher level, advanced manufacturing process control will become essential for PCB suppliers as they search for new ways to improve manufacturing processes across systems and boost overall yield.
Traceability and real-time connectivity capabilities are especially important to PCB suppliers vying for valuable government subsidies. This trend is most pronounced in Europe and also in China, where the government’s ‘Made in China 2025’ plan puts a premium value on Industry 4.0 enablement. Traceability and advanced analytics applications are preliminary steps in the development of AI-driven systems. These systems are in turn an essential element in the race towards a fully automated production floor — the ultimate aim of Industry 4.0 — and PCB suppliers are eager to demonstrate their competencies here.
The trend toward production digitization is accelerating among PCB suppliers with mass-market ambitions in smartphones and premium electronics. Increased momentum is inevitable as suppliers of next-generation PCBs and embedded components seek to service the fast-growing market for autonomous vehicles — a market for which the margin for device error and passenger safety is razor thin.
It’s this market that’s perhaps most likely to propagate PCB data automation requirements westward to industrialized European and North/South American factories. Passenger-carrying autonomous vehicles (and the robots that assemble them) will understandably be subject to the most stringent specifications for PCB and process integrity.
Traceability with complementary analytics tools helps to bolster assurances that the PCBs targeted for these vehicles have been readily trackable, accessible and adaptable to process controls throughout the production cycle. Down the road, real-time connectivity will help ensure that process irregularities are identified and remedied immediately after a defect is detected. This connectivity will also provide insight into machine status and related alerts and enable remote control of machinery. These capabilities are central to the Industry 4.0 vision and will be indispensable as AI takes hold in smart factory automation.
Tal Lev-Ran is an Industry 4.0 Marketing Manager at Orbotech Ltd. Joining the company in 2013, she has held several engineering roles on the automated optical inspection (AOI) product team. Prior to Orbotech, Tal held variety of positions in the semiconductors industry at Applied Materials and Numonyx. She holds an engineering degree from Ben Gurion University of the Negev.
>> This article was originally published on our sister site, EE Times.