Refusing Limits in Active Device Characterization
2021-07-07 | 12 min read
Jan Verspecht has been involved in advanced research and development in the field of metrology since he first joined Hewlett-Packard thirty years ago. He now considers his role to be an "IP Generator," thinking up ways to revolutionize how active devices and components are characterized, modeled, and tested. He holds 19 patents and “countless ideas,” each of which has a great story behind them. Since I don't have enough room on this post to tell them all, I'll share some of his philosophies with you, including my take on his inventor’s manifesto below.
Jan’s journey as an inventor is one of many stories that Keysight began sharing during National Inventors Month, in our Refusing Limits series. We’ll dive deeper into the unique approach each Keysight inventor brings to solving problems and explore how they are refusing limits to help Keysight and its customers push the boundaries of technology.
This article series looks to shine a light on how inventors push through technical limitations to develop solutions to problems that most deem impossible or intractable. I’m guessing from that smile on your face now that you welcome such challenges.
Yes! The more people say that something can’t be done, the harder I will work to find a solution. One of our patent sets here [at Keysight] was the result of trying to get an instrument, a network analyzer, to measure something that it was completely not designed to do. As the tests that we are conducting today on emerging technologies are becoming more and more complex, we need to find ways to do more with less, as they say. I wanted to take a vector network analyzer that was built to apply one tone at a time and find a way that it could handle a lot of tones, up to a million. Everyone told me that it couldn’t be done. It was impossible. Perfect—my kind of challenge!
While others might have been focused on the hardware, I was thinking about the math.
While others might have been focused on the hardware, I was thinking about the math… developing a software algorithm that would overcome the hardware limitations. The trick was in preserving the accuracy of the measurements with no hardware protections in place. Solving this tough problem required going back to basics and I ended up using prime number theory. I assembled an amazing team of colleagues, and it took us all about two weeks to stand up a demonstration of our innovative method. Over time, that simple concept fueled half a dozen patents and invention disclosures that branched out from the initial Method and Apparatus for Spectral Stitching Using Reference Channel and Pilot Tones. It was a great example of how one disruptive and original invention could be applied in novel ways to enrich an array of product offerings.
Yes, I see that you brought your colleagues with you to this discussion, Troels Nielsen and Augustine Stav. I take it that you see your work as a shoulder-to-shoulder collaboration, yes?
Absolutely. We're at the point now where Troels, Augie, and I can sit down over an evening of home-brewed beer and leave with two or three patentable ideas in our pockets. Our team is a great example of the value of diverse thinking to such a grassroots effort. Having a master`s degree in physics, Augie had little knowledge of electronics, but he learned fast, is unbelievably creative, loves problem-solving, and he has a knack for turning novel ideas into tangible deliverables. Troels brought domain strength in nonlinear large-signal network analysis and EVM (error vector magnitude) measurement across multiple high-tech industries, hoping that our approach would have implications across a number of test and measurement devices.
We are a team that pushes our instrumentation to the edges of what the hardware can deliver.
Collaboration is key as we stretch the thinking on experiment designs, trying to forge an optimal experiment design that will extract information about a device in the most efficient and effective way. By that I mean people just assume that a signal generator is perfect. Or they just look at the output signal over a device, but that doesn't really tell you what the device is doing, because the output will also depend on the input. So, we need to collectively consider the whole problem as we think through an optimal experiment design. That's how we came up with what we like to call compact test signals that are really optimal test signals to extract just what you need, that can run up to 1,000 times faster without a lot of overhead.
What’s the biggest challenge you face when presenting a new idea?
When I first started out, it took me a couple of years before any of my ideas materialized into a product. When things are working well and your company is successful, it can be hard sometimes to challenge your legacy. Incremental thinking can be easier to sell but it's the big new ideas that get you out ahead of the market needs. All those business journals encourage us to look for ways to disrupt the status quo, but the hard part is that you have to disrupt yourselves first.
Inventors at a large company can rely on legal teams to help them navigate the disclosure and patent filing waters. What does that process look like from your perspective?
I actually filed my first patent during a period when I broke off for a while as an independent consultant twenty years ago. I was a legal team of one at that time and had to learn the process myself, which has informed my work ever since. I ended up spending twenty bucks on some do-it-yourself patenting handbook which allowed me to understand the process enough to have an informed discussion with an examiner at the United States Patent and Trademark Office in Alexandria [Virginia, USA]. With that guidance, I was able to start writing my own patents without an attorney, and now, with that knowledge tucked away, I’m able to craft patent applications with a minimal amount of time-consuming iterations. A little research goes a long way.
What was the most important takeaway from your patent process research?
Document everything. You need a paper trail of your research on existing IP—that’s the first step with any problem—figuring out whether anyone else has addressed this problem already. Documentation is also a way to foster collaboration and knowledge sharing. It helps to pique interest across various research communities, articulates the problem that you're trying to solve, and tracks your progress. You're all on the same page—literally. And because we sometimes find ourselves way ahead of a market need, you need to keep all your creativity well documented, as an idea that was fast-failed five or ten years earlier might have commercial potential today.
By documenting the particular problem that you are trying to solve, you are also codifying the clear goal you have in mind, as a north star to keep you from straying too far as you search and explore potential solutions. The invention process reminds me of “slime mold,” this brainless, single-celled organism that can learn and problem solve as it hunts for richer food sources. It probes every option, but “remembers” all the dead ends. Invention can be an organic process as well, but as our deep and complex ideas grow and grow, we need to write them down!
An Inventor’s Manifesto:
- Inventions are not isolated events. They grow like a tree of evolution.
- Our value comes from simplifying things, not from making things more complicated.
- Your goal should be to find ways to obsolete your own products; over the long-term, that’s how you help your company to survive and thrive.
- Have empathy that your invention might upend someone else’s entire life’s work.
- People have to unlearn—it is often much harder to convince an expert of the value of what you’re doing than a novice.
- Break down your belief system and analyze a problem at its foundation (down to its basics, back to prime number theory if you have to, and then build it back up again).
- Innovation is less about process and much more about progressing towards an objective.
- Invention isn’t just a matter of leaping from one hilltop to a higher one; you have to navigate the valleys in between and find ways to climb the technical cliffs.
- Labeling a great idea as intellectual property makes people want to hold onto it for dear life.
- Knowledge sharing isn’t something that you can easily put into numbers, but we know that the trust it builds is invaluable.