My Years With General Motors

by Alfred P. Sloan Jr.

  The dramatic accomplishments of science in recent years have captured the imagination of everyone, and have caused industry as a whole to move into a "research era." The word "research" is used in industry in a number of different ways: to denote scientific discovery, or advanced engineering, or even traditional and routine product improvements, the last being clearly an abuse of the term. Research has always been difficult to define in a way that distinguishes the more basic or fundamental type from applied research. There is no commonly agreed hard-and-fast line as to how "basic" a thing has to be, objectively, to be called "basic research." The definition upon which there seems to be wide agreement is that basic research is the pursuit of knowledge for its own sake. In this sense, we in this country are not doing nearly enough.

  The solution to this problem lies mainly in the universities and in government activity, but in recent years the question has arisen of the role of private industry. Obviously the major portion of the work must be undertaken in the universities. They have the academic viewpoint, the purpose, the tradition, the atmosphere, and the talent for seeking knowledge for its own sake. My personal viewpoint is expressed in the Alfred P. Sloan Foundation, which supports a program of basic research in the physical sciences in the universities. That this research is basic is indicated by the fact that the foundation bets not on the project but on the talent of the individual, who selects his own research in accordance with his individual interests, desires, and ability.

  Obviously, too, basic research that requires unique and expensive facilities beyond the resources of universities is properly the province of government establishments such as the Bureau of Standards and the more recent Atomic Energy Commission and the National Aeronautics and Space Administration.

  As to the participation of industry in basic research, the question has two parts: research inside the industrial organization and that done outside the organization but financed by it. I think, first of all, that since the outcome of basic research is the foundation of the knowledge used in industry, it is appropriate and an expression of enlightened self-interest for industry to make outside grants to universities for basic research. In other words, industry should do this because in the long run it will help industry. I think that shareholders and management will agree in principle with my position on this.

  The extent to which industry should engage in basic research inside its own borders is a complex and somewhat unsettled problem. I cannot see how industry in its own work can properly distract its attention, in a large way, from its own practical projects. From the standpoint that basic research is the seeking of knowledge for its own sake, it is apparent that it does not belong, in a primary sense, in industry.

  It does not follow, however, that industry should not engage at all in basic research. To a certain extent I think it should. A compromise is necessary. The scientists seek knowledge primarily for its own sake; industry seeks knowledge for eventual application. It is, however, reasonable for industry to engage in basic research in specified areas where any advance in knowledge, however speculative, is likely to be of eventual use to the industry: a kind of scientific reconnaissance. In other words, industry might legitimately employ scientists to work on basic research within the industry in areas where the scientists' fields of interest coincide with those of the industry, even though the motivations of each are different.

  For example, a scientist might say: "My chief interest is the relation of the properties of individual metals to the properties of alloys. I don't care of what use this is. I want to know why it is." A producer of alloys could hardly help but be interested in the results of the research. So long as the motives of the scientist and the industry are not prejudiced, it is reasonable for them to establish a working relationship. The compromise lies not in motivation, but in the overlapping objective fields of interest. The scientist's "basic research" may be the industry's "exploratory research." This is the kind of basic research I think industry is justified in engaging in, since there is a reasonable expectation of an application regardless of the scientist's disinterested motive. To avoid any possible limitation on research activities, we need side by side both the industrial and the academic approaches.

  In sum, therefore, my argument is this: that basic research, defined as the search for knowledge for its own sake, belongs primarily in the universities; that industry should give support to basic research in universities; and that industry has a special interest in engaging in several types of basic research inside of industry where there is a common broad area of subject matter. Useful results come more quickly from basic research than heretofore and so a basic research group inside industry becomes a valuable intelligence group in the physical sciences. And the presence of scientists, well known for their work in basic science, helps morale and the prestige of the industrial laboratory and of the enterprise itself.

  The Engineering Staff

  The Engineering Staff provides an intermediate, medium-range link between the Research Laboratories and divisional engineering activities. It chiefly develops new engineering concepts and designs, and appraises them for commercial application.

  We did not have a department or section of the corporation under the title "Engineering Staff" until 1931. But the various persons and functions that were brought together to make up this staff already existed. Some of them went back to the early twenties. When, for example, Mr. Hunt and Mr. Crane worked up the new Pontiac car in the Chevrolet Division in 1924 and 1925, that was a species of improvised staff operation for a special purpose. The General Technical Committee, formed in 1923, was another step toward an engineering-staff concept. Our divisions then differed greatly in their engineering practices, and in the quality of their engineering work. Some of our products were well designed; others were not. I have described the lack then of any extensive interchange of information among the divisions, or any means which insured that this would take place, and how the General Technical Committee, uniting research, divisional engineers, and general executives, was made to serve this purpose. The General Technical Committee, growing as it did out of our experience with the copper-cooled engine, was the beginning of all engineering co-ordination in General Motors. From this committee came the corporation's first regular testing program. Cars then were being tested on public roads, and there was no easy way of telling whether the test driver had pulled up at the side of a road, taken a nap, and then driven faster than the test schedule called for to make up the necessary mileage. Once one of our engineers discovered a test car jacked up outside a dance hall with the engine running up the required mileage on the odometer.

  The most important step we took to standardize and improve test procedures was the establishment in 1924 of the General Motors Proving Ground, the first of its kind in the automobile industry. The thought was that we would have a large area, properly protected, and entirely closed to the public. It would be provided with roads of various types representing all the various demands on the motorcar from the standpoints of high speed, hills of various grades, smooth roads, rough roads, ability of a car to move through water—which is frequently required in severe storms—and the like. There we would be able to prove out our cars under controlled conditions both before and after production, and we could also make comprehensive tests on competitive cars.

  The idea was approved and the necessary capital made available. The next problem was to find out where such a proving ground could be located. What we wanted was a varied terrain centrally situated with relation to our manufacturing operations in Lansing, Flint, Pontiac, and Detroit. Michigan is rather flat, and at first we had difficulty locating an area of sufficient size that would give us all the various grades we needed. However, almost every foot of the United States has been measured topographically, and the record was available in Washington. We went to Washington and from the Geological Survey maps available there we determined a location that appeared to fulfill our needs. Then the general executives and engineers of the various divi
sions and myself spent a day at the prospective site. We walked all over the place, ate a picnic lunch under the trees, and finally came to the conclusion that that particular area of 1125 acres—now 4010 acres—at Milford, Michigan, would meet the requirements we had in mind.

  I delegated one of my executive assistants, W. J. Davidson, to take responsibility for developing the Proving Ground, and he appointed F. M. Holden as the first resident manager. Not long afterward Mr. Holden went to Oakland at his own request and was succeeded at the Proving Ground by O. T. ("Pop") Kreusser. All three of these men contributed greatly to the success of this project.

  The land was surveyed; the straightaways were laid out so that we could check the effects of different winds on speed; a track was built and banked so that it was reasonably safe to operate cars at speeds up to 100 miles an hour or more. Engineering buildings were erected, so that indoor tests could be made in correlation with outdoor tests. Headquarters and facilities were provided for the corporation's engineers. Separate engineering headquarters and garage facilities were eventually provided for the staffs of the engineering departments of the various divisions, so that they could preserve their divisional autonomy in testing. Chevrolet, for example, could do its own testing if desired, in addition to that being done by the corporation. A clubhouse was erected that provided sleeping quarters, dining facilities, and the like for those attached to the Proving Ground operations, since the Proving Ground itself was a considerable number of miles from any town where commissary facilities were available.

  In those days I used to spend a day and a night, sometimes longer, at the Proving Ground every other week. I would go over the engineering of General Motors' cars and competitive cars. I would examine what was being done in the way of testing future products. The Proving Ground thus afforded my associates and myself a wonderful opportunity to find out what was going on in the automobile industry from the engineering point of view. To the original Proving Ground we have since added a special, desert proving ground at Mesa, Arizona, and a station to test cars in mountain driving and a garage and shop facility to service our test cars at Manitou Springs (Pike's Peak), Colorado.

  The General Technical Committee, it will be recalled, acted in the 1920s as a kind of board of directors for the Proving Ground as a part of its work in co-ordinating and standardizing engineering procedures throughout the corporation. It also administered certain other central staff activities, such as the Patent Section, the New Devices Section, which evaluated technical devices submitted to the corporation by outside persons, and a foreign engineering liaison section.

  But the General Technical Committee had no engineering staff of its own. Advanced engineering of corporation-wide interest was conducted in the 1920s either by the Research Laboratories or by the engineering departments in the individual operating divisions. We made a practice after a few years of having each operating division undertake some problem of long-range significance. These divisional engineering departments of the 1920s were the ancestors of the modern corporate Engineering Staff. They were not the best arrangements in the world, for the divisional responsibility is to the product the division is sponsoring. The division, charged with bringing out a new model every year, constantly encounters new problems which are its primary responsibility. When you inject a piece of long-range research and development into this situation you are superimposing on an already loaded organization something to which it cannot properly give its attention. Recognition of this led to the formation of the Engineering Staff which was responsible to the central office.

  This great advance in the engineering area was begun in 1929 when O. E. Hunt of Chevrolet was made the corporation vice president for engineering. Mr. Hunt then succeeded me as chairman of the General Technical Committee and took on the task of coordinating the advanced engineering work of the whole corporation. Under Mr. Hunt's guidance, the advanced engineering in the divisions became a corporation staff responsibility. The functions of the old General Technical Committee were gradually absorbed into other parts of the corporation. Special product-study groups, for example, were developed for certain major problems. The product study group was a "task force" of engineers assigned to a specific mission. Although in most cases situated physically within a specific division, a product-study group was a corporation activity, financed by its own corporation budget. The top operating group would try to identify the major directions in which car development was moving. We would then locate a capable engineer and set up a group under him to work on a selected problem. We set up the first product-study group in 1929 to adapt the Chevrolet for the use of Vauxhall in England; this group also designed cars for Opel in Germany, and other small cars. Afterward we set up the Suspension Product Study Group and the Transmission Product Study Group in the Cadillac Division (subsequently involving the Oldsmobile Division and the GMC Truck & Coach Division), and the Engine Product Study Group in the Buick Division. The first was responsible for developing independent front-wheel suspension; the second developed the fully automatic Hydra-Matic transmission for passenger cars and related units for larger commercial vehicles; the third was responsible for many improvements in the car engine. As time passed we changed the product-study groups from corporation task forces situated physically within the operating divisions into permanent separate organizations engaged in the continual process of research and testing in four vital areas—power development, transmission development, structure and suspension development, and the design of new types of cars. Eventually we took them out of the divisions and brought them together in the Engineering Staff, and called them development groups. They form the heart of the Engineering Staff today.

  The Engineering Staff is closely linked to the Engineering Policy Group through the vice president of engineering, who directs the Engineering Staff and is chairman of the Engineering Policy Group. Since this group reviews the major steps in the development of new models and major departures from current engineering practice, it is in close touch with the engineering work of the operating divisions. The best thought of the Engineering Staff thereby makes a direct impact on the work of the divisions, and they make a direct impact on the development work of the Engineering Staff. The present organization, I believe, ensures the most rapid discovery of new concepts of engineering and their translation into current operating motorcars.

  The Manufacturing Staff

  Our over-all engineering work may logically be viewed as falling into two areas: one centers on the product and the other on the process of making it. The Manufacturing Staff works with conjectural, experimental, and pilot-model concepts; when these concepts prove successful in solving problems, they are adopted and used in our regular manufacturing operations in the form of improved manufacturing tools, equipment, and methods. This staff deals principally with the various aspects of manufacturing from the time the materials enter a plant until the finished product is shipped. These include machine and tool design, plant layout, materials handling, plant maintenance, equipment maintenance, work standards, methods engineering, materials utilization, and process and equipment development for the fabrication, final assembly, and test of the product. (Note 14-2.) This staff's general aims are to improve product quality, increase productivity, and reduce the cost of manufacture.

  Centering these activities in a single corporate staff was the idea of one of our executives, B. D. Kunkle, who in 1945 felt that there was need for the same kind of function in the manufacturing area that the product-study group fulfilled in the product-development area. The manufacture of automobiles had rapidly become a more and more involved process requiring a constant study of new materials, new machinery, and new methods. Hence the idea of specialists to develop ideas for use in the manufacturing process. Logically, this was a staff function, and as such could be better fulfilled by a corporation group than by the individual divisions.

  The technical work of the Manufacturing Staff is largely centered in the process-engineering activities of its Manu
facturing Development Section, where the problem of automation arises. Process engineering necessarily includes automatic operations. Beyond the semiautomatic and automatic machine looms the image of the semiautomatic and automatic factory—the whole vague area summed up under the term "automation," in which it is often difficult to distinguish science fiction from practical manufacturing possibilities. The Manufacturing Staff will play a large role in this field in General Motors. How far automation should go is a difficult question which will have to be decided on the highest policy levels of the corporation. General Motors and the Manufacturing Staff have tended to be somewhat more cautious in this area than some other manufacturers. There exists a widespread belief that "if it's automatic, it must be good," but our experience shows that this is not always the case.

  A good, balanced view on this subject was given in a paper presented before General Motors' 1958 Conference for Engineering and Science Educators, by Robert M. Critchfield, who was then in charge of process development. He said:

  In recent years we've all heard a lot of talk about automation. It seems to me that most of this talk has done little more than confuse a great many people, including a few in the engineering profession, as to the true implications of the word. As you know, automation is nothing new; it's merely a relatively recent word for a process that has been going on in manufacturing for more than half a century, perhaps even as far back as the time of Eli Whitney's successful attempts to mass produce muskets for the continental armies. I can recall that we had some types of transfer machines and other automatic production devices 35 years ago in General Motors, which was long before the word "automation" was coined. Our misconceptions seem to stem from the fact [that] the literature overflows with too many notions that automation is the obvious solution to the mass production of a particular part or product involving a number of highly repetitive hand operations. Nothing could be further from the truth. The decision to mechanize or not to mechanize a production process or operation involves much more than the number of repetitive operations; it involves a good many fundamentals of economics .. .