What will shape electronics manufacturing over the next three to five years? In this guest Q&A, Waldemar Christen, Senior Vice President CRM at ASMPT SMT Solutions, shares his perspective on the electronics manufacturing trends that matter most to EMS providers and the OEMs they serve.
ASMPT SMT Solutions, the technology and market leader, is recognized worldwide for its first-class, future-oriented solutions in electronics manufacturing. Its portfolio ranges from SIPLACE placement machines and DEK solder paste printers to the SPI Process Lens system and the WORKS Software Suite for planning, controlling, analyzing, and optimizing SMT production. Waldemar covers the hardware demands of AI servers, the role of data and intelligent automation on the SMT line, how to stay flexible in high-mix production, and why closer collaboration between machine suppliers, EMS providers, and OEMs builds more resilient supply chains.
If I had to single out one development, it would be the rapid expansion of AI infrastructure. Not so much because of AI itself, but because of the hardware behind it. The new generation of AI servers is creating entirely new demands for electronics manufacturing: larger and more complex PCBs, a rapidly growing number of components in a wide range of sizes, shapes and weights packed into a very limited space, and extremely high quality requirements. In addition, components are already being placed on inner layers that are subsequently integrated into the PCB. As a result, placement machines are becoming increasingly relevant in PCB manufacturing itself.
Manufacturers are collecting more data than ever before. The real challenge is to put this data to good use and to better connect machines, software and processes.
I believe five topics in particular will shape the industry over the next few years: the manufacturing of AI and high-performance computing systems, data-driven decision-making, intelligent automation, more integrated manufacturing platforms, and closed-loop systems for continuous process optimization across the SMT line.
Ultimately, however, there will not be one single technology that makes the difference. The companies that succeed will be those that bring these developments together in a meaningful way.
AI and automation are currently among the most widely discussed topics in electronics manufacturing. From a technical perspective, a great deal is already possible today. The key question is where automation also makes economic sense.
Electronics manufacturing has very different automation requirements from industries such as automotive. It requires much greater flexibility and shorter payback periods. In principle, a great deal can be automated, but not everything that is technically possible is also economically viable.
Many modern SMT solutions, including those from ASMPT, already use AI-based technologies such as machine vision, machine learning and intelligent data analytics. These technologies make processes more transparent, help identify deviations at an early stage, and support employees in making complex decisions.
Realistically, what we see today is a coexistence of people and automation. Products, components, materials and requirements are constantly changing, and particularly in high-mix production, human expertise remains essential.
For this reason, I primarily see AI as an assistance system. Its strength lies in analyzing huge volumes of data, identifying correlations across different process steps, and highlighting potential quality risks at an early stage. The companies that come out ahead will be those that are best able to combine human experience, automation and intelligent software.
We often talk about machines, but increasingly it is the data that determines how efficiently and reliably a production line operates. Modern SMT lines generate vast amounts of data. The real challenge is not collecting this data, but putting it to effective use in intelligent manufacturing.
Traceability based on production data is nothing new. It has been standard practice in our industry for many years. The key development today is the ability to process available data faster, turn it into a basis for timely decisions, and increasingly automate the decision-making itself.
We can see this, for example, in print process optimization. Based on process and SPI data, corrections can be made automatically and cleaning cycles adjusted without operator intervention. The goal is to identify trends early and take corrective action before they turn into problems.
The same applies beyond ongoing production. In service, data is increasingly enabling a predictive approach. Rather than relying solely on preventive maintenance, we can identify trends and schedule maintenance based more closely on actual needs.
Ultimately, these capabilities directly address our customers' key KPIs: higher throughput, better quality and lower total cost of ownership.
When we talk about efficiency, many people first think about the performance of individual machines. However, the greatest potential for EMS providers lies in optimizing the manufacturing operation as a whole.
One key lever is the use of high-performance lines. If, for example, two existing lines can be replaced by one higher-performance line, this not only reduces the required floor space but also significantly cuts the effort and resources associated with operating each additional line.
At the same time, manufacturers do not necessarily have to rebuild their entire production setup. The SIPLACE V can be easily integrated into existing lines, allowing companies to benefit from the higher performance of the new placement platform without having to invest in a completely new line. Existing hardware and software solutions can continue to be used.
One of the most common sources of inefficiency is focusing too much on individual machines and not enough on the line as a whole. A machine may perform extremely well on its own, but if there is a bottleneck elsewhere in the line, that performance cannot be fully utilized. What ultimately matters is the performance of the entire line.
Another important factor is standardization. Clear, consistent processes provide the foundation for stable and efficient production. In practice, however, we often see workarounds being introduced as soon as problems arise. That may be necessary in the short term, but it should never become a permanent solution.
The key is to identify and eliminate the root cause of a problem. If a workaround is necessary, it should always be temporary, with a clear objective: to return to a stable, standardized process.
Flexibility has become a key competitive factor for EMS providers. Customers expect shorter lead times, smaller lot sizes and an ever-growing variety of products.
Flexibility starts with powerful line software. The WORKS Software Suite, for example, can be used to plan and optimize production processes. WORKS Planning also helps determine which products can be efficiently grouped into setup families and when a dedicated setup is the better option. The goal is to avoid unnecessary performance losses and make the best possible use of available line capacity.
Equally important are extremely short changeover times. With small lot sizes and frequent product changes, a changeover needs to be completed within minutes. The less productive time is lost during changeovers, the more flexibly an EMS provider can respond to changing requirements.
The production technology itself must also support this flexibility. The SIPLACE V placement platform can be easily integrated into existing lines. The SIPLACE V L adds further options for material supply, including more feeder slots and a space-saving, easily exchangeable SIPLACE V Tray Unit. At the same time, many aspects of the modular platform are already familiar to operators, which keeps additional training requirements to a minimum.
The key is to ensure that flexibility does not come at the expense of productivity. Manufacturers producing a wide variety of products in small lot sizes need to master both: fast changeovers and maintaining high line performance.
One of the biggest changes we are seeing today is in the automotive sector. The shift toward electric mobility is fundamentally transforming the industry and, with it, the requirements for electronics manufacturing. At the same time, AI is increasingly finding its way into vehicles. In the future, virtually every vehicle will have a powerful server board capable of processing large volumes of data quickly and enabling AI-based functions.
The global expansion of AI infrastructure is another major driver for electronics manufacturing. Investments in data centers and high-performance computing systems are increasing demand for ever more complex PCBs and creating new challenges for manufacturers.
We are also seeing many companies diversify their supply chains and manufacturing locations. Supply security, resilience and geopolitical risks play a much greater role today than they did just a few years ago.
I believe the relationship between OEMs and EMS providers will become even more collaborative and strategic in the coming years.
This is not simply about working more closely together. EMS providers can increasingly become technology enablers for their OEM customers. New manufacturing technologies are opening up possibilities that can play an important role as early as the product design stage.
One example is the direct placement of active semiconductor dies from the wafer, as enabled by the SIPLACE CA2. The cost per placement is increasingly approaching that of placing conventional SMT components, opening up new possibilities in product design.
The relationship between OEMs and EMS providers is therefore increasingly evolving from a traditional customer-supplier relationship into a technology partnership.
Supply chain challenges have changed considerably in recent years. Today, I see strong market fluctuations as the biggest challenge. A year or two ago, the market was relatively weak, while today we are once again seeing very strong demand in some areas. Managing these fluctuations is a major challenge for companies.
EMS providers therefore need to be able to adapt their production quickly to changing demand and capacity utilization.
At the same time, this also creates opportunities. Companies that are able to manage these fluctuations and adjust their capacity accordingly will gain a competitive advantage.
The closer machine suppliers, EMS providers and OEMs work together, the earlier they can identify changes and prepare for them. Especially in times of strong market fluctuations, having this information early is critical.
Open and continuous communication is equally important. The more transparently information about expected demand and potential changes along the supply chain is shared, the better all parties can align their capacity and planning.
We see this in our own business as well. Our account managers work very closely with our customers, giving us early insight into expected demand. This allows us to plan our own production not only on the basis of existing orders, but also on forecasts. It helps us prepare for changes in good time and ensure our ability to deliver.
For me, this is a key part of a resilient supply chain: transparency, early communication and close collaboration across the entire value chain.
In my view, digitalization plays a key role in making supply chains more robust. It is less about digitalization for its own sake and more about making information available faster and using it to make better decisions
One major benefit is connectivity across the entire supply chain. Digitalization provides a common communication platform that allows OEMs, EMS providers and machine suppliers to coordinate quickly and directly.
Ultimately, it is not the amount of data available that matters, but what you do with it. Digitalization provides the foundation for faster, better-informed and more forward-looking decisions.
The biggest challenge is the increasing complexity of PCB placement. Higher component densities, larger and heavier components, and ever tighter tolerances leave significantly less room for process deviations. At the same time, the value of the products and components being processed is rising considerably.
The goal has to be 100% yield across the entire process chain. It is not enough to have a defect-free PCB coming off the line at the end. No high-value component can be lost during the manufacturing process either. When individual components can be worth as much as US$30,000, every single error can quickly become very costly.
The quality requirements are at least as high as those in the automotive industry. This works to our advantage at ASMPT, as we have many years of experience with similarly demanding quality requirements in automotive manufacturing.
As a result, it is no longer enough to inspect quality only at the end of production. Quality has to be ensured throughout the entire process. Solutions such as WORKS Optimization monitor the entire SMT line, helping to identify process deviations at an early stage and ensure quality across the individual process steps.
ASMPT WORKS Software Suite used to plan and optimise SMT production
The first step is to look at the process as a whole. New technologies can make a significant difference, but they cannot replace stable processes. In solder paste printing, for example, stencil design is already a critical factor. If the fundamentals are not right, even high-quality solder paste and optimally adjusted process parameters cannot fully compensate.
The major advantage of new technologies is their ability to capture and analyze process information very quickly and make it available in a way that supports the right decisions. This is exactly where solutions such as WORKS Optimization come in. Edge technology enables large volumes of data to be processed directly and relevant process information to be made available quickly.
Today, many decisions are still made by people. As we gain more experience with these technologies and the systems become more robust, more of these decisions will be automated.
It is important not to optimize individual process steps in isolation. Electronics manufacturing is more complex than it may appear at first glance. Focusing on one aspect while ignoring the interdependencies can easily lead to optimizing in the wrong place. What matters is always the process as a whole.
According to Waldemar Christen of ASMPT, the rapid expansion of AI infrastructure. AI servers need larger, more complex PCBs, with more components packed into less space and extremely high quality requirements. He also points to data-driven decision-making, intelligent automation, more integrated manufacturing platforms, and closed-loop process optimisation across the SMT line.
No. Waldemar sees AI primarily as an assistance system. It analyses large volumes of data, identifies correlations across process steps, and highlights quality risks early. In high-mix production, where products and requirements change constantly, human expertise remains essential.
In optimising the manufacturing operation as a whole rather than individual machines. A fast machine can't deliver its full performance if there is a bottleneck elsewhere in the line, but replacing two existing lines with one higher-performance line can reduce floor space and operating effort.
By combining line software that groups products into efficient setup families, changeovers completed within minutes, and placement equipment that integrates with existing lines. Additionally, the production technology itself must also support this flexibility. The aim is flexibility that doesn't come at the expense of line performance.
By sharing information about expected demand and potential changes early and transparently. The closer the three work together, the sooner each can align capacity and planning, which matters most in times of strong market fluctuation.