News and New Products

Q&A: Bob Lucky

Hard work makes its own luck.

By Bill Schweber, Executive Editor -- EDN, 8/4/2005

Robert "Bob" Lucky, PhD, is the former head of the legendary Bell Labs. Bell was the home of much of the vital research that our industry now employs, including the transistor. It also was the birthplace of some pure research having no apparent application, such as discovering that the omnipresent background noise of the universe is due to remnants of the big-bang theory. While at Bell, Lucky also played a key role in the development of adaptive equalizers' filtering strategy, a key component in all high-speed modems. He's also the co-author of the seminal textbook Principles of Data Communications and a long-running columnist in IEEE Spectrum. In addition, he has just been named chairman of The Marconi Society.

What led you to become an engineer?

When I was in high school, I didn't have a clue what engineers did. I was good in math and physics, and I liked to build radios and electronic equipment, so people said I should be an engineer. I marvel at how clueless I was. Now, after a long career as an engineer, I'm still not sure what they do.

I'll never forget my first week at college. I had to take mechanical drawing, and the instructor told us that all engineers had to start their jobs on the drawing board. That actually wasn't true even then, but I believed it-especially because I wasn't much good at mechanical drawing. During a test, I drew something badly, and, in exasperation, I threw my compass down on the drawing. It bounced off the drawing and flew out the open window next to me. I don't remember how I did on that test, but it couldn't have been good.

How did you get into writing your IEEE Spectrum column and other less-technical items?

People sometimes ask me how they can get their own column, usually hinting that they could do better than I do. Sometimes, I tell them that you just have to be "lucky." In the early years of my column, I used to worry that someone would take it away from me. People would send me sample columns that they had written and copy the Spectrum editor. That hasn't happened now for a long time, and, after 23 years of writing the column, I don't really worry about things like that any more.

I owe the column to the Spectrum editor of 1981, who was Don Christensen. He had asked me to do a book review of Tracy Kidder's The Soul of a New Machine—still a great book, by the way—for the institute's newsletter. They liked my review and asked me to write a movie review of Tron. Then, I wrote a review of Michael Crichton's Congo for Spectrum. I was quite abashed when Crichton himself sent a rebuttal to my review to Spectrum. I had said that his technology was faulty, and his rejoinder was that his book was fiction, and he could make up stuff. I agreed with this statement but said that the fact that he had included lots of references from IEEE journals gave the reader the idea that the technology was correct. In private communication, we both agreed that the other had a point. Moreover, I have to say that his subsequent novels all are based on great technical ideas.

After these reviews, Christensen suggested that I try a column and see how it went. The rest is history. It is truly a great privilege to have a column, and I often take a moment to appreciate the luck and the honor.

In the beginning, I had to submit my columns for clearance through the Bell Labs review process. I was embarrassed to do this, and, apparently, so were the reviewers. Then the public-relations department told me that maybe this clearance wasn't necessary-that I was on my own for these columns. Now, wherever I go, someone will come up to me and say that they like my columns. I never tire of hearing this, and, whatever work I do on them is worthwhile. People often tell me that I write well "for an engineer." I'm never sure whether this is a compliment.

Did you think your work at Bell Labs would have such a broad impact on digital communications?

Actually, I've never thought that my work did have that kind of impact. I was fortunate to be in the right place at the right time, 1961, when modems were first being developed. I'm sure that, if I hadn't invented the adaptive equalizer, someone else would have done it a month later. Much of technology is like that. There are, of course, exceptions, and the great work of [the late] Claude Shannon, [the "father of information theory"], is the biggest exception to that general rule.

Any comments on the state of engineering and science education?

People say it's bad, particularly math education in the secondary schools, but I haven't been involved in that issue. I have been concerned, however, with what an engineer should learn in college. I'm not sure I have any answers, but I do believe that, in college, you need to learn how to learn, and I'm not sure that this is a real focus. Everything I studied has been obsolete for decades. All that remains is a general engineering culture, some mathematical principles, and an ability to learn new things as time goes along.

How has the industry changed, for better and worse, since the 1960s and 1970s?

Old Bell Labs people like me shake their heads at what has happened and congratulate themselves on having enjoyed the “golden years” during the ’60s and ’70s. No one then worried about corporate profits. It was all about innovation and having the best possible network. That best-network thing was like a religion. I tried to explain this mindset to [the late] Judge Harold Greene in the AT&T antitrust trial, but I think I failed. After all, look at what happened!

The decline in corporate support for research is a big problem for the United States. With incremental development moving offshore, we need to run with innovation to keep in place, and we’re not doing that. It is especially bad in telecom, where the operating companies have abandoned their research labs, and Bell Labs itself has undergone five splits: Bellcore, Lucent, AT&T, Agere, and Avaya.

Today, universities are authoring about 85% of telecom-research publications. That’s where research is now being done. That isn’t all bad, but my theory is that industry research, pointing out the significant problem areas, has to gently lead the academic research. Of course, the telecom-research problem is only part of a bigger picture. I believe that corporate executives have lost their faith in the return on investment in research. It’s hard to prove, and competition is more aggressive now than it was 30 years ago. In telecom, for example, companies like MCI proved that you don’t need research labs to succeed financially. Or maybe they didn’t, as it turned out.

Are there any technological developments that you were relatively certain would happen, but didn’t, such as videophones? What about the reverse?

I never thought that videophones would be a winner. I had one of the AT&T Picturephones in my office. The only person who ever called was my boss, and I had to stare into the thing. Finally, I think I had the last one in the world. There was no one to call. Today, anyone can have a videophone on the Internet, and it’s cheap, but it’s still not a big deal. It turns out that the information is in the speech, and video doesn’t add that much. In some scenarios, it could even detract; that’s my theory, anyway.

Like many technologists, my problem in retrospect has been not in knowing what would happen, but believing that it would happen much sooner that it did. I started a group working on programmable modems in the late 1970s: too early. I started a group working on computer games over telephone lines in 1976: too early. My biggest failing, which I believe I share with about every other engineer, was not fully understanding or believing in Moore’s Law. I remember in 1977 giving a talk to the head of switching development advocating a digital switch. But digital is more expensive, they said, and everything you can do digitally, we can do with analog. I should have shown them a curve of Moore’s Law. But, alas, I didn’t.

Where do you see the next area of likely major advance?
 
I haven’t the foggiest idea. The big new ideas come out of left field. Powerful processors, speech recognition, and so forth are givens. I’d like to see some kind of communication using quantum phenomena and Einstein’s “spooky” coupling between paired particles—something really mysterious and mindblowing.

Editor's note: The above is an expanded version of the article that appears in the print edition. The PDF below reproduces the printed version.