CHAPTER IV Features of Technology Development On Being the Right Size In the preceding chapters we discussed the history of research institutions and some of their accomplishments. It is important to recognize that the institutions with which we are dealing are not new and have, in fact, evolved over some three hundred years. The merit of the historical approach is in demonstrating the provisional, time-bound nature of any institution organized to do fundamental or applied research. What we will attempt in the remainder of this book is to analyze in detail how contemporary research and development institutions are actually managed, and to suggest how the management can be improved. We touched earlier on the achievements of modern research institutions. Nuclear technology, space technology, and electronic technology such as transistors and integrated circuits have been developed in laboratories of the kind that are the subject of this book. It might also be worthwhile to talk about some failures. Paradoxically, one cannot say that there have been any failures in the development of a particular technology once its physical principles became known. But while technology developments tend to be successful, certain technology development institutions have failed, especially in the Federal sector. Failures in the commercial sector are also known, but in that case we have measurable failures of products rather than of laboratories. In many cases, fully-developed products were brought to the marketplace and for one reason or another price, inefficiency, supersession by more efficient competing technologies — turned out to be commercial disasters. Some of the best-known of these failed technologies include the rotary engine, Corfam, a synthetic material developed by DuPont, the electronic facsimile transmission system developed by Xerox, and fluidics, a technology using liquids or gases to perform functions ordinarily performed by electronic devices (ref. 55). Despite intensive marketing research and advertising campaigns, these products or technologies were things for which there was either no demand or not enough to justify mass production. Failure also means the closing of a specific institution after the objective for which it was first organized was achieved.* A good example is the U.S. Naval Radiological Defense Laboratory once located at the Naval Shipyard in San Francisco. Its mission was to apply modern methods of nuclear safety technology to decontaminating naval vessels subject to nuclear attacks. The laboratory was established in 1946 to 1947, and during the next decade achieved most of its goals; decontamination techniques for ships now exist that are probably adequate for what needs to be done. Why, then was the laboratory shut down? Probably because management recognized too late that its mission was completed and that the organization had to find new problems sufficiently important to justify its continued existence. Either there were no such problems or they were not pursued vigorously enough. The laboratory finally closed in 1968 on the ground that there simply was nothing left for it to do. The laboratory had made its contribution to resolving an important technical problem. That was not enough to save it. In the private sector, technology development laboratories also occasionally fail. During the 1960s, the Northrop Corporation felt that it had an important future in space technology. Accordingly, a corporate space laboratory was established. Unfortunately, the timing of the step taken by Northrop was not very auspicious. The company's space laboratory was started just as spending on space technology by the Federal Government was beginning to decline. Despite the quality of the people Northrop hired, the management of the company finally concluded that the space laboratory could not be sustained and closed the institution. Let us take the analysis a step further. Why are some laboratories closed once their missions are accomplished and why do others remain * Some reviewers wondered why a laboratory which completed its mission should be considered a failure. They would argue that it is not failure, but success, if an agency successfully solves a problem or carries out an agency mission, even if in so doing it works itself out of a job. For two reasons we disagreed. The first is that it is inherently inefficient to create a separate institution, staff it with scientists and engineers, and spend perhaps several millions of dollars on equipment, only to shut it down at a certain point. The same objection would hold for a contractor-operated facility, even if the cost to the government were partially concealed. We concede that special circumstances might justify such an arrangement, although the creation of a new division within an existing laboratory might be less costly. The fundamental objection to regarding the completion of a mission as a criterion of success is that, in such cases, "mission" is construed too narrowly. A laboratory set up to develop one kind of catalytic converter or guidance system could probably make contributions in other aspects of pollution control or navigation technology. By closing the facility after its original mission - or rather, assignment is completed, the corporate or government sponsor forecloses the possibility of building on the experience gained. Successful laboratories are able, again and again, to reinterpret their missions in light of changing conditions and that is really what we mean by success. open? What is it, for example, about a laboratory such as the NASA Langley Research Center or the U.S. Naval Research Laboratory that makes them productive organizations long after their original reasons for being have been forgotten? Both of these institutions are in their seventh decade and yet they have managed to retain their vitality. What, in fact, are the correct ways of evaluating the performance of a technology development institution? One way might be to use past achievements and then make small extrapolations from them. A laboratory that can successfully produce nuclear warheads capable of working from ballistic missiles, another that can develop reentry systems for the Apollo program, and a third that can develop the swept wing principle for high-speed subsonic aircraft probably will continue to make important contributions. But such an argument is often not enough. The authorities within the agency who are in charge of preparing budgets will ask, “What have you done for me lately?" A research facility cannot survive on its record of achievement, by serving as a job shop for other agencies, or even by modest departures from its original mission. Successful research and development laboratories share certain features that are apparently independent of the particular technology they are pursuing. Three seem to be particularly important: 1. Cost. The cost of operating a technology development institution is (in 1983 dollars) about $75 000-$100 000 per employee. This number, multiplied by the number of employees, is the institution's budget, and it tends to hold good independently of what the laboratory does or whether it is public or private. The probable explanation is simply that research and development is labor intensive, and that the dollar figure cited equals salary plus overhead, plus some funding for equipment needed to carry out technology development. 2. Professional and Support Personnel. The ratio of direct program people to support people in technology development institutions is between one to one and one to two. On the average, in order to keep one person busy in a technology development task, it takes one support person meaning a secretary, technician, librarian, and the like - who does not work directly for a single scientist or engineer. Those institutions that are more test-and development-oriented tend to have more support people for each direct professional than those oriented toward basic research. 3. Size. Almost all permanent research and development institutions range in size between 1 000 and 7 000 people. There are very few operations that are smaller and very few that are larger. The reason for this range of sizes is something like this: The lower limit is determined by the fact that if the institution is too small, there will be too little flexibility for a few people to strike out into new territory, or for new ideas to spill over into research work. Thus the U.S. Naval Radiological Defense Laboratory mentioned earlier had a total staff of about 800. There were not enough groups of two or three or four people delving into areas unconnected with the laboratory's current mission but that might lead to new missions. Institutions seemingly must have more than about 1 000 people before the kind of flexibility that makes for the institution's survival exists. The upper limit in size is determined by the difficulty, for management, of staying intellectually on top of an institution having more than 6 000 to 7 000 people. (Los Alamos and Lawrence Livermore Laboratories, with just over 7 000 employees apiece and budgets of $421 and $515 million, respectively, are near the upper limit. The Sandia National Laboratory, operated by Western Electric for the Department of Energy, has 8 000 employees at three locations and a budget of $738 million (ref. 56).) Although laboratory directors should not attempt management of research in detail, they must nevertheless understand thoroughly the objectives of work in progress. This is difficult in the largest laboratories; hence a certain creaking of institutional joints, resulting from too many communication channels and not enough feedback. The existence of organizations like Bell Labs, with 19 000 employees, does not disprove this; such organizations are best thought of as federations of semi-autonomous installations of 1 000 to 5 000 persons each.* Problems of Research Diversification In the Federal sector, every research installation is always "under judgment" by a variety of groups. The performance of each institution is reviewed annually for budget purposes at the very least. However, there are also a great many other reviews, ranging from those performed by committees of the National Academy of Sciences or Engineering under contract to the agency being reviewed, to those carried out by the General Accounting Office. When a technology development institution gets into real trouble, there are usually a number of special reviews before a decision is finally reached to close it or to make significant changes. Nevertheless, changes do happen fairly rapidly in the Federal technology development establishment. In 1968, the year the Naval Radiological Defense Laboratory closed, the Defense Department also: • Consolidated several Air Force activities with elements of the Army, Navy, and NASA into the Eastern Test Range (Cape Canaveral, Florida), the Western Test Range (Vandenberg Air Force Base, California), and the White Sands Missile Range in New Mexico; * As a result of the divestiture of AT&T at the end of 1983, some 3 000 of Bell Labs' 22 000 employees were transferred to Bell Communications Research, owned and operated by and for the divested telephone companies. |