The Changing Face of Industrial Research
W. DALE COMPTON
West Lafayette, Indiana
Giving this talk is a great pleasure—not only because I get to meet all of you and be part of this meeting, but also because this venue brings back many memories for me.
You are meeting at the Ford Research Laboratories during tumultuous times for the U.S. automotive industry. Although I can’t offer insights into the current state of the industry or Ford, I believe I can offer some insight into the factors that have led to the creation, and then the near demise, of some of this country’s great industrial research establishments.
Major changes have occurred in recent years. Bell Labs is a distant memory. The size and focus of IBM research, GE research, Westinghouse research, and Xerox research have all been reduced, and research activities in many other companies have undergone similar changes. The focus of the industrial research that remains has changed—except for the pharmaceutical and biomedical industries, industrial research no longer includes basic research.
Why and how did these changes come about? Let’s begin by examining the philosophical justifications that made industrial research popular some 50 years ago. This philosophy laid the groundwork for the great tide of industrial research that ultimately shaped the research posture today.
The story begins with World War II, when scientists and engineers from all over the country were called upon to leave their universities and come together to create the technologies and systems that were crucial to winning the war. The key developments included radar, the proximity fuse, the atomic bomb, and many others.
In November 1944, following the cessation of hostilities, President Franklin Roosevelt wrote to Dr. Vannevar Bush requesting his recommendations for post-
war activities. In Roosevelt’s words, “New frontiers of the mind are before us, and if they are pioneered with the same vision, boldness, and drive with which we waged this war, we can create a fuller and more fruitful employment and a fuller and more fruitful life.” Dr. Bush, who had been the director of the Office of Scientific Research and Development during the war, responded with a report, Science—The Endless Frontier.
A number of very important actions were taken in response to the recommendations in that report. First, in 1945, the Office of Naval Research was founded, the first government agency responsible for funding research that did not necessarily address immediate requirements for the military. This was the start of federal funding for basic research in our universities. In January 1946, the Research Grants Office was created at NIH to administer projects of the Office of Scientific Research and Development and to operate a program of extramural research grants and fellowship awards. A few years later, in 1950, the National Science Foundation (NSF) was created. NSF is still the principal source of funding for basic research in the physical sciences, engineering, and social sciences. Other federal departments soon followed suit and established their own extramural funding activities. Many of you have participated in one or more of these federal programs. In fact, NSF is providing partial support for this conference.
A number of actions were taken by industry soon after this. Principally, industrial research was initiated—primarily in the physical sciences and engineering. Back in 1951, Ford created the Ford Research Laboratories, as Gerhard Schmidt1 mentioned earlier. During the 50th anniversary celebration for the laboratories, Ford characterized those 50 years in a very useful way. The period from 1951 to 1970 was called the “golden era” of research, 1970 to 1985 the era of redirection and regulation, and 1985 to 2001 the era of relevance. If we replaced regulation in the middle era with deregulation, as in the case of telecommunications, airlines, and energy, this characterization would be rather accurate for industry as a whole.
THE GOLDEN ERA
What was the emphasis during the golden era of research, the first 25 years after World War II? At that time, people looked at what scientists and engineers had done during the war and concluded that they could do the same for industry. The principal justification for supporting research was that good research would lead to good products and good profits. Please note the absence of any mention of the topics to be explored. The second justification was people. A research
organization staffed by outstanding individuals would be available to consult on internal company problems and would serve as the eyes and ears of the company in the global world of science and technology.
Based on these two justifications, a wide variety of corporate laboratories flourished. Great progress was made in science, publications were abundant, and opportunities for employment abounded. Again, notice that the research output was not directly tied to the profits of the companies.
The quality of science in physics, chemistry, and metallurgy at Ford from 1951 to 1970 was outstanding. If one were to compare the physics department at Ford with academic physics departments throughout the country, I think it is safe to say Ford would have been in the top ten—maybe even the top five.
Al Overhauser at Ford was the discoverer of the Overhauser effect, which has been of enormous importance for solid-state physics. The first SQUID—any of you in magnetic measurements will recognize the SQUID as the most sensitive detector of magnetic fields that has ever been developed—was also invented at Ford. The first frequency doubler for the laser was demonstrated here. The scientists at Ford were widely known and recognized as outstanding. Scientists in the Ford chemistry and metallurgy departments were similarly talented.
THE ERA OF REDIRECTION AND REGULATION
Now we enter the era of redirection and regulation—1970 to 1986. Allow me to share briefly some of my findings on joining Ford in 1970. First, the people were outstanding. But, although they had great eyes and ears in the technical world, they had no effective communication with the operations sector of the company. In fact, they had little credibility with operations. Programs in the physics area were related to general relativity and solving math problems that were fun but had hardly any relevance to the company’s operations. More important, I found no systems effort at all in the laboratory.
During this era, we made serious efforts to refocus the laboratory on problems relevant to the company while keeping the research as long range and basic as possible. At the time, the company was beset by external demands that required new directions for research. The Clean Air Act became law and emissions regulations, fuel-economy regulations, and regulations limiting emissions from manufacturing plants were passed.
It was not hard to find research areas that could provide a fundamental understanding that would be relevant to meeting these demands, but it took time for people to reorient their thinking and embrace problems that were relevant to the new needs of the company. In due course, however, a number of new activities were initiated.
An atmospheric-science program was started, and a major effort in systems engineering was undertaken. These soon began to provide dividends. Ford was the first company to develop a simulation model for hybrid vehicles and then to
demonstrate the viability of a hybrid. Ford developed the first flexible-fuel vehicles that could operate on 100 percent methanol, 100 percent gasoline, or any mixture of the two—all automatically. We developed the first electronic engine controls in 1972. In 1975, they went into production as the first fully programmable electronic, adaptive control system in a production vehicle.
That production program led to the largest off-campus reeducation of engineers up to that time. Most of the mechanical engineers involved in the development of the engine and the controls needed up-to-date training on digital electronics, so many of them pursued master’s degree programs in electrical engineering at Wayne State University. General courses were taught on the Wayne State campus. Courses based on proprietary information were taught at Ford.
Catalytic converters were critical to controlling emissions from vehicles. Ford sponsored the development of the monolithic catalyst structure, which later became the model for the industry. The key active ingredients in the catalyst were platinum and rodium, and reducing the amount of platinum became a longstanding goal. Haren Gandhi,2 who is sitting here in the front row, participated in that effort. You may have noticed, on the outside wall in the large hall, the Medal of Technology that President Bush presented to Haren for his efforts to improve the efficiency of catalysts, and, hence, reduce their cost by reducing the amount of platinum they required.
Despite the success of the catalyst, it was also an unfortunate example of how hard it is to communicate across “silos” in a large company. While Haren was successfully reducing the amount of platinum in each catalyst, and thus reducing the cost of each unit, the financial sector of the company held a very large forward position in platinum. Because there was no good mechanism for discussions between the company’s research and financial sectors, Ford lost a lot of money when the need for less platinum was announced and the value of platinum plummeted.
Research in many other areas was also ongoing—sensors, stamping dies, new paints, new high-strength alloys, and so on. Basic research also continued, at a significantly lower level but at a high enough level to be effective. In other words, there was still an effective mass concentrated in areas that would logically support the large, more applied programs that, in turn, supported the operating divisions. I am sure Gerhard will understand that maintaining basic research programs was possible only because we were able to hide them. The programs directly related to the company’s operating divisions were large enough and important enough that management was not interested in asking questions about the other programs.
The new regulatory environment had a profound impact on the company,
which found itself working toward technical objectives with which it had no experience and little knowledge. Communication by researchers with bright people in operations who were willing to look at the needs of the company was possible, but not easy. There were still many detractors—people at high levels in the company who wanted to reduce the research budget by as much as 30 percent in one year. Fortunately, clearer heads prevailed, and the research laboratories were given enough time to make the necessary reorientation. If they had been forced to do so in a very short time, it is unlikely that they would have continued to exist.
THE ERA OF RELEVANCE
Now to the era of relevance, namely post-1986. In this era, essentially all programs must contribute either to the products or the processes used by the company that sponsors them. This is true not just at Ford, but throughout industry. As companies have reduced the size, or even eliminated, their research laboratories or, at least, eliminated long-term research in their laboratories, many have increased their cooperation with universities, particularly in the bioscience and engineering sectors. As a result, little basic research is being done today by industry, although many industry segments continue to support some basic research, mostly in universities rather than in their own laboratories.
But there are some problems with this arrangement. For example, graduates interested in pursuing careers in research have fewer opportunities. In addition, questions have arisen about ownership of intellectual property and, most important, about the funding philosophy of industry and federal agencies supporting research in universities. For example, how long a view can researchers take? And how many risks?
Let me share with you my personal biases about what we have learned in the last 60 years. Good research can be done on relevant problems, but those problems frequently lead to questions that can only be answered by fundamental research, which takes time. The reason the management of research in industry is so very difficult is that researchers must identify and be working on problems well before the operations sectors even realize they have a problem. Those problems cannot be solved immediately. Research must be out in front, and that takes tremendous foresight, which, in turn, requires that the research sector be in close touch with operations. That’s the only way these problems can be anticipated and understood. Only after that, can a company decide what can be done to solve them and which problems will only be solvable through basic research.
A research organization in industry is in a very fragile position. If it is too close to operations, the pressures for short-term results may increase to the point
that long-term research is crowded out. If operations are held at arm’s length, however, the research sector can lose its credibility and the support of the people who would benefit the most from its results.
Those of you who work in universities may think none of this applies to you. I hope that this is not the case, because choosing the right problem to work on at the right time is as critical to your success as to the success in industry. In fact, it is critical that you get funding to pursue that research.
THE PROBLEMS AHEAD
Just as we learned that research in industry can only prosper in the long term when the research sector maintains contact with its customer, namely the company, so must we identify and tackle the really important problems confronting not only the company, but also the country and world markets. Some of these problems are technical, and some are not. Just for illustration, I will give you examples of each.
First, a nontechnical problem—the lack of understanding of the consequences of political decisions, both local and national, related to technology. A search of the Congressional Research Office database for congressmen with “engineer” in their titles turned up only one. There may be a few more, but only one was found in that search. That is pathetic, but it reflects how difficult it is for technical people to reach out and be part of the political system. We desperately need to think about how to break down that barrier.
We must address the whole issue of technical literacy, for both technical and nontechnical people. I might observe that one of the benefits of a symposium like this is that it increases your technical literacy in subjects that are not in your special area of expertise. At my own institution, Purdue University, there are no courses that teach technology to nontechnical students. This is a travesty.
Now for the technical problems. A lot of things could be used as examples, but I will show my biases with two of them—energy independence and health care delivery. First, energy independence. We have to find alternative fuels. We have to find a way to become less dependent on the petroleum sources in this world. The conversion of cellulose to liquid fuel and coal to liquid are viable sources of liquid-based fuels, but the technology is not yet at a point that would make these viable.
My second example is a newly emerging research area for engineering. The National Academy of Engineering and the Institute of Medicine recently published a study on the subject of engineering and health care delivery. The focus was not just on bioengineering and biomedical engineering, as important as they are, but also on the system by which care is provided to people—such as system optimization, sensors, remote communications, telemedicine, and making every hospital room an intensive care unit. Another question is how we can take ad-
vantage of the Internet in the delivery of health care, the long-term role of which we cannot predict.
At the beginning of this new century, the technical community is facing two major challenges. The first is ensuring the continuing availability of innovation, which is critical to our national prosperity. The second is supporting the necessary level of research to ensure that innovation continues.
There is a tendency to think we have come full circle since 1944, from research is golden, to research is unnecessary, to a realization that research is critical to future innovation. However, we live in a time of globalization, when competition is fierce, money is limited, and expenditures that are not directly relevant to a company’s mission must be justified. In fact, this is a more difficult environment than the environment of the 1970s when we were suffering the effects of the oil embargo. We must find ways to meet these challenges through both technical and nontechnical means.
I close with a quote from the recent National Academies study, Rising Above the Gathering Storm. “This nation must prepare with great urgency to preserve its strategic and economic security. Because other nations have, and probably will continue to have, the competitive advantage of a low-wage structure, the United States must compete by optimizing its knowledge-based resources, particularly in science and technology, and by sustaining the most fertile environment for new and revitalized industries and the well-paying jobs they bring.”
I leave you with a big question. How will, or can, our institutions respond to these challenges?