National Instruments (Austin, TX) reached two milestones this year: 30 years as a company and 20 years of supplying its industrystandard measurement software, LabVIEW. Microwaves & RF spoke recently with Dr. James Truchard, president, CEO, and co-founder of National Instruments (www.ni.com) about his beginnings with the company and where measurement technology was heading in the next few years.
MRF: It has been 30 years since beginning National Instruments and 20 years since the introduction of LabVIEW. What were the early days like?
Truchard: I'd like to talk a little bit about what it was like before we started the company. I was working at the University of Texas; I had gotten my Ph.D. in 1974. I'd been work full time primarily on projects for the US Navy. We were building computer-based test systems; in hindsight, we could call these" Virtual Instrumentation" systems. The last one had three PDP11s running UNIX and had to be operated by shipyard mechanics. So it was pretty extreme at both ends. We used digitizers and software to build the instrumentation. So this really established an idea that we would bring forward for using ever more powerful computers to build instrumentation systems.
MRF: Do you remember the performance of the digitizers?
Truchard: These were state-of-the-art digitizers at that time. These were 2-MHz, 12-b analogtodigital converters (ADCs) that took about 35 W power and cost us around $5000 a piece. Now you could get the equivalent capability for a few dollars.It was the most advanced system at that time. And it really inspired us to see what could be done with computers. In starting the company, we really took a step backwards from those kinds of systems to simpler applications, simply interfacing instruments to computers. A new standard had come out in 1975. So in starting the company in 1976 we embraced the standard because we had seen how difficult it was to connect instruments to computers. The first one we did took two man-years of effort just to talk to the instrument. It was a very complex interface that needed a custom board design and custom software to make it work. So we appreciated the need and the utility of having a standardized interface that was the IEEE 488 as our starting point. And it was a good starting point because it was right between computers and instruments. What better place to be if you wanted to revolutionize instrumentation?
So we moonlighted to start those first interfaces starting in 1976 and by 1979 we were actually a profitable company, self-financing the company with very little money. That proved to be good business. From that point, we went full time and started thinking of the bigger picture, the application. We had seen how successful spreadsheets had become for solving the problem of financial analysis on a computer. Before having a spreadsheet, it might have taken you six months to right a program where the columns lined up just right. With a spreadsheet, you could do it in minutes. So we hoped to do a similar thing in developing LabVIEW.
Starting in the early 1980s, Jeff and I worked to find an answer. I wanted this solution that did what the spreadsheet did; Jeff wanted to invent a programming language. We both got our way with LabVIEW. It was introduced in 1986 and really was revolutionary for its time. No one had really made a practical, graphical programming language with data flow like we did with LabVIEW. And it turned out to be a very good model approach for instrumentation. Typically, we had observed when we started doing measurements in our previous applications; we would start with a block diagram. That was the natural language for engineers to communicate with each other. And, obviously, instruments had front panels. So, if you could combine those two elements to create an application development environment, you could really make a much better and more user-friendly programming environmentfor instrumentation. So Jeff did that. To this day, no one has found a better way to solve that problem than we did with LabVIEW.
MRF: What were people using before that time?
Truchard: They were basically using the functionality of the instruments that the vendor defined and use the simple calculators at the time to capture results or generate a plot. It was very simple use of the instrumentation that didn't really provide advanced analysis. With LabVIEW, you have your choice. You can either control traditional instruments, or you could build your own instruments with user-defined functionality. And I had worked previously in the area of sonar where there weren't any offthe-shelf instruments to do our work. That's why we had to build these dedicated digitized instruments. So we saw a need for broad use of instrumentation that could be defined by the customer rather than have it packaged in the manner of traditional instruments.
MRF: Isn't it true that your early work in that sonar instrumentation has developed into what you now call Virtual Instrumentation (VI)?
Truchard: The basic concepts and the use of a computer to build the instrumentation rather than the fixed functionality of the traditional instrument, it's the concept that brought us to the ideas that we would use. Of course, the technology that we came up with, and Jeff's computer-science background gave him a good start on the solution.
MRF: Is Jeff still involved with the company?
Truchard: Jeff is very actively involved with the company, even going on some recent press tours. But he is busy adding capabilities to Lab-VIEW. The capability that he is adding now is very much related to the microwave and RF area. We are really extending LabVIEW's capabilities in that area. He is busy with the basics of LabVIEW to make multiarray programming even easier.
MRF: LabVIEW has been on the market for 20 years, but has not been widely used by RF and microwave engineers. Hasn't it been used by test engineering more in other areas of electronics?
Truchard: But it has been has used by RF and microwave engineers with traditional instruments. These RF and microwave instruments are typically very complex and very expensive, and because they are very complex, automation was often very good. Companies like Nokia, Ericsson, and Motorola all would use LabVIEW to program these instruments and they would have instrument drivers that would package functionality in a way to make it easier to make these very complex instruments work. These instruments were typically working around wireless standards, and there is a long list of them, and LabVIEW was the tool of choice by these wireless equipment manufacturers for testing their products.
MRF: You mention the drives from the instrumentation vendors. Doesn't that imply a kind of "buy in" from the test instrument manufacturer into LabVIEW?
Truchard: Exactly. Over the years, we have worked closely with the different instrument vendors, such as Agilent Technologies, Tektronix, and Rohde & Schwarz. For the last several years, LabVIEW has been evolving and we have literally been adding thousands of algorithms to perform RF testing directly in LabVIEW. So the standards for wireless and for RF testing now can be built in LabVIEW at much lower cost and much faster time to market. We've added hardware in PXI for RF upconversion and downconversion, and then LabVIEW technology in which we have made significant investments over the years, to do testing of base stations, cell phones, all kinds of RF products, RFID, games.
MRF: Do these algorithms ever come from the customer?
Truchard: Very definitely, and the customer can build their own. It was a customer who was one of the key sources. It was a customer who had worked at Motorola previously, on IRIDIUM project, and started his own company, and ultimately we bought his technology as a baseline. And now we have expanded on that for our modulation toolkit. We have many alliance partners who are now building dedicated test systems around key standards like Bluetooth, CDMA, and different wireless standards. They take our base technology and because of the leverage they get out of the efficiency of using LabVIEW they can very quickly build a standards-based test system at lower cost than traditional instrumentation. Because we are PC based, we also get performance advantages when we use high-performance PCs for our solutions.
MRF: Has GPIB faded over the years in its used as an instrumentation interface?
Truchard: It's been going along all these years. My son was working in the labs at the University of Texas and he was using the same antiquated instruments that I was using when I was in the labs, so these instruments stay around a long time and we see GPIB staying around a long time as well. USB is very widely used now, and Ethernet in some cases. There are some instruments now that are sold without GPIB capability, and a user has to add a converter to get the capability. In PXI, we have our PXI bus that we use, and PXI Express, which has the bandwidth directly to a low-cost PC processor and that is a significant advantage for testing. If you look at audio, almost all audio testing is now done with virtual instrumentation. And it has all moved to PC, and National Instruments has been a big benefactor of that. We feel that the next step is RF and IF digitizing to bring the data in and out of the main memory of the computer at significantly lower cost.
MRF: It appeared that the company made a concerted move into RF some years ago. Was that the case?
Truchard: We felt that RF was a significant opportunity because of the high cost of traditional instrumentation in that area. It was a real opportunity to decrease cost and decrease the time to market for the latest technologies and the latest standards. We recognized that there was a possibility for us to make many of the RF and microwave measurements in these emerging communications standards.
MRF: Where does National Instruments stand on LXI?
Truchard: We actually have what we call a 1588 controller that we can use to control these traditional instruments, and we already have several hundred drivers that work with Ethernet. We like to talk about multiple standards. For distance, Ethernet is the best bus. For close-in systems, USB is a better choice, it is faster and easier to work with. And it is supported very heavily by the PC. Obviously, most people use PCs now. Even the CAD stations have moved to PCs. We believe that a multiple of standards will be around for a while. People still want to use their old instruments, so GPIB will be around. USB is the easiest to work with. Ethernet for distance applications. We already have our own Ethernet products, like Compact RIO. We are already supporting Ethernet, but we see it as one of a multitude of standards. In terms of virtual instrumentation, we have been doing virtual instrumentation for two decades now, and really have delivered a lot of value to our customers.
MRF: National Instruments has been recognized for seven years in a row by a leading business publication as one of the top 100 companies in the United States to work for. Are there any secrets for other managers?
Truchard: We are one of 13 high-tech companies on that list; it used to be a lot more high-tech companies on the list. We are proud of our employees in creating that environment. We have worked hard over the years to create a good environment for engineers. When I started the company, I used to say that I wanted to create a job that I liked. And that has been literally true. And one of the aspects of doing that was creating jobs that other people would like. If they weren't having fun then it would be hard for me to have fun. So we really worked on that. As a small company, it was more of the walking-around management and meet everyone or if someone seemed a little down you could step in and ask them what the problem was. As we grew, we kind of outgrew the ability to do that. We dedicate a lot of energy to training on how to learn how to create a very good environment. Now we even have leadership training where key managers are in a program with 11 four-hour sessions that I co-teach. We wanted to have the senior management bring everybody on board to learn about virtual instrumentation, to learn about the philosophy of the company, to really get excited about the vision and mission that we have in the company.
MRF: When you started the company in your garage, did you ever imagine that it would grow to its current size, about 3800 people worldwide?
Truchard: I have a "canned" answer to that in that I always wanted the company to grow. So, theoretically the company had to reach this size some time or another. I had worked at a lab that didn't grow, and it was painful. We would hire young graduates and then they would leave, because there weren't any career opportunities for them. So, in starting National Instruments, I wanted to have a company that would provide really solid career opportunities for our employees. And that meant we had to grow. We couldn't just stay the same size and provide those opportunities. That has been one of the founding mantras for the company, to create that environment where they could not only have fun at work, but also have a long-term career path.
MRF: Has managing that growth been a problem?
Truchard: With a small company, you struggle with financing— selffinancing in our case. As you grow larger, the environment and the finding of business that allows you to continue to grow become important. I like a challenge, and I like risk, and dealing with it. I like to take an idea that I think is very theoretically and carry it to tens and thousands of people who are using it. When I was working at the university, I might work on a research project for two years, and at the end of that time maybe five people would understand it. That wasn't enough for me. I like to see a lot of people get value out of what I was doing. We are reaching a lot of people now, in vehicle testing, games, entertainment systems.
MRF: Are you still involved with product design and development?
Truchard: I spend much of my time on product strategy. I've been heavily involved in the RF testing products, for example. We have an experimental program with universities now for RF design where I have been very hands on and I have met with all the professors. I have been very involved at extending LabVIEW into different areas, including RF and microwave communications. I am also involved in management and investor relations. But I really enjoy the product design and development and that is what helps the company grow.
MRF: Can you talk about what you offer in the area of military testing?
Truchard: Our roots go way back in understanding what the military is looking for in testing. We've been involved with LabWindows, LabVIEW, TestStand is very heavily being used by the military, our PXI instrumentation is being standardized by many military agencies. We've got a long track record in the military test area. We are involved with most military programs around the globe, with companies like Boeing, Lockheed Martin, and Northrop Grumman.
See Dr. James Truchard.