» F R O M T H E E D I T O R

6 IEEE CONTROL SYSTEMS MAGAZINE » JUNE 2006

How many scientific discoveriescan you think of that were madeby mistake or serendipity? With-

out checking Google, I can think ofpenicillin, Teflon, nickel titanium, X rays, and the big bang echo. Most ofus wouldn’t include serendipity as akey step in our next project, but wewould be delighted if such an eventwere to occur.

Compared to scientists, engineersseem to have fewer of these excitingand romantic tales of unexpectedbreakthroughs. When Black built hisfeedback amplifiers, he quickly dis-covered that stability was an impor-tant issue, leading to elegant analysisby Nyquist. Likewise, I suppose thatKalman didn’t expect to find a Riccatiequation at the end of his filter deriva-tion. But neither of these “discover-ies” has the instant serendipity thatmakes headlines and legends.

What is unexpected in engineeringis the way in which engineering inno-vations are used. It’s easy to think ofexamples. The magnetron was devel-oped for radar but now cooks food inmost kitchens. The iPOD was devel-oped for music and is now used forPODcasting lectures (presumablyreducing live classroom attendance).Although both uses were unexpected,unanticipated applications are notnew. The windmill governor wasadapted to the steam engine gover-nor, while amplifier feedbackadvanced from triodes to semicon-ductors without missing a beat.

But not all outcomes meet thedeveloper’s expectations, especiallywhen engineers invent with highideals, believing that the fruits oftheir labor will contribute to peace

Surprise

A snow sculpture at the Harbin snow festival.

The Harbin ice festival.

JAS

MIN

E D

UR

RA

NI

AN

NA

JUNE 2006 « IEEE CONTROL SYSTEMS MAGAZINE 7

and well-being. Philo T. Farnsworth,the inventor of television, felt thatway about the medium he invented.His innovation was a coated glasstube, scanned in electronic bous-trephedon, which he envisioned atage 14 while plowing a potato fieldin Iowa. But later in l ife heexpressed disappointment at howtelevision steals time from readingand conversation. Today, televisionis practically a metaphor for deca-

dence (for a graphic depiction, seethe movie Avalon).

Farnsworth, who died in 1971,could not have foreseen how his CRTwould develop as an essential compo-nent of computer technology. Like-wise, we continue to be surprised bythe innovative and unexpected bene-fits of the Web. We can take satisfac-tion in its many remarkable uses,except possibly the widespreadgames, gambling, and pornography.

Engineers develop technology andhand it off to others who use it inunexpected ways. Some of these usesplease us, and others do not. But wehope that what we create will benefitpeace and well-being in the long run,serendipitously or not.

Dennis S. BernsteinEditor-in-Chief

IEEE Control Systems Magazine

At the entrance to the Forbidden City. Inside the Forbidden City.

Hidden Leverage

When a process engineer attempts to construct a set of algorithms that will be the basis for automat-ing some portion of the production process, he or she first interviews those individuals who cur-

rently perform the tasks that will be automated. The process engineer must learn the detail of theiractions in order to translate their practice into the terms of a mathematical model. The algorithms insuch a model explicate, rationalize, and institutionalize know-how. In the course of these interviews, theprocess engineer is likely to run up against the limits of implicit knowledge. A worker may performcompetently yet be unable to communicate the structure of his or her actions. As one engineer discov-ered: There are operators who can run the paper machine with tremendous efficiency, but they cannotdescribe to you how they do it. They have built-in actions and senses that they are not aware of. One opera-tion required pulling two levers simultaneously, and they were not conscious of the fact that they werepulling two levers. They said they were pulling one. The operators run the mill, but they don’t understandhow. There are operators who know exactly what to do, but they cannot tell you how they do it.

—S. Zuboff, In the Age of the Smart Machine: The Future of Work and Power, Basic Books, 1988, p. 60.

ZH

EN

G M

INJI

E

HE

WE

I