What Scares Me Today in PCB and System Design

Estimated reading time: 6 minutes

After more than 30 years in PCB design, I have seen our industry evolve from relatively straightforward circuit boards to highly complex, high-density, high-speed electronic systems that push the limits of physics, materials, manufacturing processes, and engineering capabilities. I have worked across multiple design disciplines, tackled difficult technical challenges, and watched the tools and technologies available to engineers advance dramatically.

So, what scares me today in PCB and system design? It isn't smaller geometries, faster edge rates, more layers, or even the relentless push toward higher density and greater functionality. Those are difficult engineering challenges, but they are challenges we can address through sound engineering principles, experience, collaboration, and the right technology.

What concerns me most is the growing disconnect between engineering intent, manufacturing reality, supply chain availability, and the decisions organizations make under pressure to deliver products faster. We are becoming even more capable of designing extraordinarily sophisticated electronics, yet we are not always equally capable of ensuring those designs can be manufactured, sourced, validated, and sustained reliably. That is where I believe some of our industry's greatest risks are hiding.

Keeping Pace with Complexity
PCB design is no longer simply about connecting components. It requires managing electromagnetic behavior, signal and power integrity, thermal performance, mechanical constraints, materials, fabrication tolerances, assembly processes, and testability—often simultaneously. HDI and UHDI technologies, advanced packaging, finer geometries, and sophisticated material systems are expanding what is possible. But every step forward can narrow our manufacturing margins and increase the consequences of a design decision made without fully understanding its downstream implications.

Consider a high-speed design where a stackup is selected primarily to achieve electrical performance. If the board material is difficult to procure, the fabricator lacks the necessary process capability, or the required construction cannot be maintained consistently across production lots, the design may satisfy its simulation requirements yet struggle to become a repeatable product.

This is not a theoretical concern. Industry reporting throughout 2026 has highlighted increasing pressure on advanced PCB materials, manufacturing capacity, and lead times as AI infrastructure demand expands. These findings were echoed during many discussions I had while walking the PCB West conference floor. The lesson is that a design is not successful simply because it works electrically. It must also be manufacturable, procurable, testable, and reliable in the real world.

A Dangerous Assumption: Manufacturing Will Figure It Out
One of my greatest concerns is that design and manufacturing are still treated as separate phases rather than interconnected engineering responsibilities. Too often, manufacturing engineering becomes deeply involved only after the design is reportedly completed. But yet, a fabricator identifies an issue with the stackup, via structure, material selection, copper distribution, tolerances, or fabrication process. The resulting changes can trigger additional engineering work, verification and schedule delays, and unexpected costs.

Sometimes the changes are minor. Other times, they affect electrical performance, reliability, or the assumptions underlying the original design. We should not be surprised when this happens if manufacturing expertise was absent when the critical decisions were made.

Design for manufacturability, assembly, test, and reliability cannot be treated as a final checklist, or worse, an afterthought. These disciplines must influence the design from the beginning, with fabrication and assembly partners contributing their process knowledge while changes are still manageable.

The same applies to manufacturing data. Gerber data, along with additional support files, can still be used, but understand that it still requires manual effort. When intelligent product information is reduced to disconnected files, drawings, and interpretations, downstream teams may have to reconstruct engineering intent. A complete, validated digital handoff using intelligent standards-based formats such as ODB++ or IPC-2581 can help preserve that intent and reduce unnecessary interpretation. The real opportunity is not merely to transfer data faster; it is to preserve the engineering intelligence behind the data throughout the product lifecycle.

Supply Chain Resilience Starts at the Designer's Desk
Supply chain disruption is another concern that experienced engineers cannot afford to ignore. Component availability, semiconductor allocation, specialty laminates, copper foil, glass reinforcement, and geopolitical uncertainty can all influence whether a product can be built on schedule. Industry concentration around advanced computing technologies adds another layer of competition for critical resources. Yet procurement and supply chain decisions are frequently made after engineering has committed to components, land patterns (footprints), board dimensions, stackups, and routing architectures.

What happens when a critical component becomes unavailable? Is there a qualified alternative, or does the substitution force a land pattern (footprint) change, alter power requirements, introduce new signal integrity considerations, or invalidate previous verification? A component that appears equivalent on a bill of materials may not be equivalent electrically, thermally, mechanically, or from a long-term availability standpoint.

Supply chain resilience must therefore become part of engineering strategy, not simply a purchasing responsibility. That means considering lifecycle status, sourcing risk, approved alternatives, material availability, and the consequences of substitutions before a design is released. It also means recognizing that the cheapest or most readily available option today may create greater cost and risk over the product's lifetime. We cannot engineer supply chain risk out of existence, but we can design products that are better prepared to withstand it.

AI: Acceleration Without Understanding?
AI is rapidly entering the engineering conversation, including PCB layout, optimization, verification, and design automation. The potential is significant. AI can help engineers explore alternatives, automate repetitive tasks, identify patterns, and evaluate design options that might otherwise require substantial manual effort. But I have a concern: Are we becoming so focused on what AI can generate that we risk overlooking what an engineer must understand?

A tool can propose a routing solution, recommend a component, or optimize a layout against defined objectives. That does not automatically mean the result is physically sound, manufacturable within a particular supplier's capabilities, or appropriate for the product's operating environment. The quality of an automated result depends on the quality and integrity of its inputs, constraints, models, and validation. In PCB design, those inputs must reflect real electrical behavior, material properties, fabrication limits, assembly requirements, and system-level interactions.

AI should strengthen engineering judgment, not replace accountability. The engineer must remain responsible for understanding why a solution works, what assumptions support it, and what happens when those assumptions change. The greatest opportunity is to combine experienced engineering judgment with automation and AI-driven analysis. The greatest risk is confusing faster output with better engineering.

The Knowledge Gap May Be Our Most Serious Long-Term Risk
There is another issue that concerns me deeply: the loss of experienced engineering knowledge at a time when design complexity is increasing. Much of what makes an exceptional PCB designer cannot be reduced to routing rules or a design checklist. It comes from understanding the interaction between electrical behavior, physical structures, manufacturing processes, and years of lessons learned from designs that did not work as expected.

As experienced engineers retire or move into different roles, how effectively are organizations transferring that knowledge to the next generation? We need to do more than teach engineers how to operate design tools. We must develop their understanding of the underlying physics, design trade-offs, manufacturing realities, and system-level consequences of their decisions. This includes a thorough understanding of the three key competing perspectives for success in PCB design: layout solvability, performance, and manufacturing, which I refer to as the “Designer’s Triangle.”

This is especially important as automation takes on more routine work. Engineers will need to evaluate complex alternatives, interpret analysis results, manage constraints, and make decisions that cross traditional discipline boundaries. Investing in education, mentoring, structured design methodologies, and accessible engineering knowledge is not optional. It is essential to maintaining quality and reliability as the industry advances.

What Should We Do About It?
I do not believe the answer is to slow innovation or avoid complex technologies. We should embrace progress, but we must also change how we approach engineering. We need earlier collaboration between electrical, mechanical, manufacturing, test, procurement, and supply chain teams. We need constraint-driven design supported by realistic manufacturing rules and validated analysis. We need digital continuity that preserves engineering intent rather than forcing downstream teams to reconstruct it. And we need to develop engineers who understand not only how to create a design, but also how that design behaves throughout its lifecycle.

Most importantly, organizations must stop measuring success solely by how quickly a design reaches release. A release is a milestone, not proof of success. The real measure is whether the product can be built consistently, perform as intended, survive its operating environment, and remain supportable as components, materials, and suppliers change.

The Real Fear… and the Opportunity
After more than three decades in PCB design, I remain optimistic about our industry's future. The technology is extraordinary, the engineering tools are more capable than ever, and the opportunities to innovate are immense. But I am concerned that our ability to create complexity is advancing faster than our ability to manage it across the entire product lifecycle.

The next generation of electronic systems will not be successful because a single engineer, tool, or organization solves every problem. Success will depend on how effectively the entire engineering ecosystem works together, from initial concept through design, analysis, fabrication, assembly, test, and long-term support.

What scares me today is not that we might fail to design the next generation of electronics. It is that we might design something extraordinary without doing the engineering work to ensure it can be built, delivered, and trusted. That is a risk we can address, but only if we recognize that the PCB is not an isolated design exercise. It is a critical part of a much larger system, and engineering responsibility must extend well beyond the layout.