Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Manufacturing Systems Research in the United States: An Overview Executive Summary Widespread agreement denotes that most manufacturing in- dustries in the United States are in poor condition compared to their strongest international competitors. Many U.S. manu- facturers suffer from growing cost and quality disadvantages in national and international markets. To a certain extent, these problems derive from international economic conditions outside the scope of this report. Nevertheless, this panel believes that large improvements in U.S. manufacturing competitiveness can be achieved through an increased awareness of the full scope and growing importance of manufacturing in the modern era, along with larger investments in related research and advanced educa- tion. Major elements of modern production systems are people, ma- chines, computers, and the communication links among them. An integrated systems engineering approach is essential to maximize product quality, production efficiency, and flexibility in future factories. Research is needed to provide better integration and compatibility of hardware, software, and people operating in the system. Related areas in which research is needed to support the goal of computer-integrated manufacturing (CIM) include mod- eling, simulation, control systems, and communication networks. In addition, more research is needed on certain unit processes in 216
MANUFACTURING SYSTEMS 217 manufacturing, particularly various materials-forming processes and closed-Ioop process control. In the past, government-supported research on manufactur- ing systems has been only a minor contributor to progress in this field. Recent federal initiatives via the National Science Founda- tion (NSF), including the establishment of two major Engineering Research Centers on aspects of manufacturing, are highly positive steps. We recommend that research on manufacturing systems be further strengthened through the establishment of one or more multilateral research consortia by industry, perhaps within the range defined by the Semiconductor Research Corporation (SCR) and the Microelectronics and Computer Technology Corporation (MCC). We expect to see research participation by universities, but do not anticipate major federal support for any such consor- tium. The flow of new talent out of the U.S. educational system and into manufacturing is inadequate both quantitatively and qualitatively. Although changes are under way at U.S. universities and within U.S. companies that could remedy this situation, a real gap in orientation and educational quality still exists between typical product development engineers and their counterparts in manufacturing. Furthermore, most manufacturing engineers now in practice, and even many new graduates, are ill-prepared to address electively the difficult problems to be faced on the path to true CIM systems. These weaknesses are barriers to the goal of achieving engineering integration throughout the manufacturing process from product concept through successful production. We recommend several specific actions to improve U.S. research and education in manufacturing engineering. These recommendations, set forth at the end of the report, address issues of education, technical standards, and professional engineering activity. Introduction and Backgrolln~ Discrete-product manufacturing in the modern sense of the term began with the industrial revolution.* For most of its first *This report focuses on discrete-product manufacturing, as distinct from continuous-flow processing, such as petroleum refining.
218 DIRECTIONS IN ENGINEERING }SEARCH 100 years, such manufacturing was not critically dependent on ad- vanced scientific and engineering knowledge. Fundamental limits on materials and processes were seldom pressed. Often, materi- als and energy were used wastefully, and there was little concern with environmental effects. Manufacturing commonly comprised a sequence of unit processes through which material passed with very little accompanying flow of data. Engineers and workers un- derstood their objectives largely in empirical terms, and managers usually were qualified more through experience than on the basis of scientific or engineering knowledge. Over the past 20 years, most of those circumstances have changed. Since the advent of the digital computer as a tool of industry, manufacturing has become more information-intensive. The concept of factories as integrated systems under electronic control has emerged. The possibilities for flexible (programmable) automation and for real-time optimization have become clear. Great potential benefits are evident in product quality and cost, production efficiency, flexible response to changing needs, and improved health and safety conditions. Many forces are at work to make manufacturing systems more complex in the future. Systems must be adaptable to product evolution. Such systems must also be capable of making the wide variety of products and variations that the customer demands, and of using the variety of raw materials available. Today the goal is integrated manufacturing systems that encompass the full scope of production activities from design through fabrication, assembly, testing, and delivery. This goal is not being met either rapidly enough or effectively enough by U.S. manufacturing industries. Although the basic cir- cumstances of manufacturingthe technology and the economic environment have changed worldwide, U.S. manufacturing in- dustries are finding it difficult to adapt to those changes. In many engineering fields (e.g., computers, aerospace design, communica- tions), U.S. industry is noted for effective application of modern scientific and engineering knowledge. Unfortunately, this Is not true in manufacturing. Nowhere near the full benefits of modern technology and knowledge have been realized in the U.S. man- ufacturing environment. In important product areas (e.g., steel, automobiles, consumer electronics), foreign competitors are the
MANUFACTURING SYSTEMS 219 leaders in developing and exploiting advanced manufacturing ca- pabilities.* Among the reasons for the United States' deficiency in this area are current policies and practices of industry, govern- ment, and universities. This deficiency cannot be sustained. Manufacturing is highly significant for the nation's welfare. The U.S. manufacturing in- dustry generates approximately 24 percent of the gross national product and about 65 percent of the tangible goods produced. The nation is unlikely to achieve full employment without a strong, in- ternationally competitive manufacturing capability. Steadily im- proving productivity in the manufacturing sector is essential to our national economic well-being. Furthermore, public expecta- tions for environmental quality can be achieved only by applying sophisticated controls to manufacturing plants. This panel has evaluated the overall direction and strength of engineering research and educational programs that bear on U.S. capabilities in manufacturing. In this report we suggest im- provements for existing programs and propose some new research- related activities involving industry, universities, and government. We also recommend actions intended to upgrade the practice and profession of manufacturing generally. Although there may not be complete agreement with some of the proposals made by the pane! to strengthen the nation's industrial competitiveness, we have en- countered widespread support for the need for more high-quaTity research and teaching in manufacturing in order to achieve this goal. Our proposals are intended to produce that result. Research and education are inherently Tong-term activities. The panel has not attempted to describe the important shorter range planning, development, and implementation activities that must be undertaken by manufacturers seeking to improve their competitive position. The Manufacturing Research Agenda The pane} believes that there are major deficiencies in research aimed at manufacturing technology. The underlying knowledge *See, for example, all of the reports listed in the Bibliography under the National Academy of Engineering, National Research Council, and Office of Technology Assessment.
220 DIRECTIONS IN ENGINEERING RESEARCH base is still quite limited and highly empirical. Furthermore, the flow of knowledge from research to application needs strengthen- ing. Reasons for these national deficiencies are described in later sections of this report; most can be traced to limitations in the training and education of personnel. In this section we address some examples of critical areas of manufacturing research. We found it useful to classify manufacturing technology in three cate- gories, each of which needs increased research support. SYSTEMS INTEGRATION Systems integration is the most critical area. Basic under- standing in this area falls far short of meeting the need for sys- tematic, generic approaches to the design of CIM systems. Major programs of interdisciplinary research are needed to provide better engineering methods for systems integration. Manufacturing systems are complex, and enormous volumes of data are required to describe and control them. In traditional factories the "data" used are embedded in the (fixed and variable) configurations of the machines, in drawings and other documents, and in the minds of the human operators. Data translations among these elements of manufacturing systems have been performed by humans, with corresponding limitations on the maximum amount of data and the speed and accuracy of its manipulation. Although some excellent manufacturing systems do not make intensive use of computers, these systems usually depend on long-term continuity and excellence of (human) engineering support. As manufactur- ing processes and end products become more complex, and as demands for flexibility increase, it is clear that computers will be an increasingly necessary element of integrated manufactur- ing systems. The term computer-integrated manufacturing implies that all relevant data are available throughout a network of com- puters, so that they can be used as needed to achieve desired overall results for the complete production system. Improvements in product quality, manufacturing flexibility and efficiency, and human productivity will be achieved through more intensive use of information. Figure 1 illustrates the scope of CIM. The benefits of CIM have already been amply demonstrated. Portions of this technology have been installed in several factories in this country ant] abroad, and have produced reductions in: tooling costs, parts inventories, control and scheduling problems,
221 ~ 0 , 0 CL U) ~ =0 ~ ~ s ~ ~ ~ a 1 ~ 1 ~ 0 E E 0 - ° ~ ° 0 o V] U) a) U2 U) no o - o v 0 ' tn C, ~ ~ ~ ~ o ' ~ ~ o o ~ ~ ° ~ to 0 ~ C' o ~ 2 ~ N so ~ ~ ~ =_ I O ' ii. ~ ~ L - o C) o o C) 1 1 1 to, o v cr a at at . To v L. Do . L. 4. o ID D. ., 0 , ~ 0 o fir ~ 0 o 0 o 0 1 1 ~ · ~ ! -
222 DIRECTIONS IN ENGINEERING RESEARCH materials handling requirements, start-up and rework, engineering changes, defective work, and unit costs; along with improvements in: manufacturability, utilization, throughput, and quality. CIM has proven difficult to implement. One reason is in- compatibilities among data formats, hardware, and software of computers and production equipment made by different vendors. Generally, suppliers have tended to develop unique proprietary data formats, hardware, and software in order to induce customers to purchase all components from one source or to offer selective performance advantages. Integration of the people and technologies of CIM is one of the most important (yet most difficult to plan and manage) as- pects of automation. This problem extends from the factory floor through engineering, design, purchasing, marketing, finance, and management. Needed research in this area must involve a vari- ety of engineering and nonengineering specialists. A system-level approach must be taken in designing elements of CIM. Research should be aimed at providing new hardware and software that is modular, compatible with other systems, adaptable to new re- quirements, and easily understood by the people who will use it. The goal of CIM is to improve productivity and quality in pros auction. Even when CIM is achieved, formal optimization (in the mathematical sense) of a complete production process probably will not be practical because of the large number of variables and many nonlinear, dynamic, and stochastic relationships. However, engineering strategies based on breaking down complex systems into hierarchies of simpler components have the potential to be practical and helpful. The hierarchy for manufacturing likely will consist of models and controls at several levels for example, the individual machine level, the multimachine process cell level, the factory level, and the multiplant level. A long-term goal is to develop the techniques of artificial in- telligence to provide computer-based capabilities such as inference and intuition hi ways that can be applied to manufacturing. Such developments would be of great value in achieving full automation and in optimizing the production process, and might also be ex- ploited in training humans for their roles in design and production.
MANUFACTURING SYSTEMS 223 MODELING, SIMULATION, CONTROL, NETWORKS Important areas still exist in which available models of materi- als, physical objects, and manufacturing processes are inadequate for the needs of CIM. Modeling of products, fully describing their geometric and other characteristics and functions, is a prerequisite for manufacturing systems integration. Current models for solid objects do not provide unambiguous geometric information, and they provide almost no information on nongeometric characteris- tics and functions. Research is needed to Seine structured computer data bases for product modeling. Computer-aided design should generate the original data base for a desired product. The goal is to auto- matically derive, define, and verify the complete data base for planning, controlling, and implementing production and testing of a product directly from this data base. Expert systems techniques may be useful in resolving difficulties encountered in this process. Another long-term goal is to provide computer-aided design sys- tems with automatic feedback of information about the production implications of design decisions. (This link is labeled "Cost and Capabilities" in Figure 1.) Such a capability has a strong potential for improving product producibility and quality and for reducing costs. Continuing research is needed on computer modem for manu- facluring processes. Although some processes are quite well mod- eled today, in other cases existing models are inadequate to predict the sensitivity of final results to process variables. Models includ- ing dynamics are needed to understand and control such effects as too! chatter in metal cutting and precision pattern alignment in microelectronics. Computer simulation techniques are still evolving rapidly, driven by advances in hardware, simulation software, and process models. High-quality graphics provided with relatively inexpen- sive workstations surely will be widely exploited. Such develop- ments are being integrated into engineering curricula; graduates with simulation and graphics experience will be eager to exploit these tools. Robotics and programmable automation have attracted the at- tention of many researchers interested in manufacturing automa- tion. Both anthropomorphic robots, which are humanTike in some
224 DIRECTIONS IN ENGINEERING RESEARCH respects, and nonanthropomorphic robots are important. Physi- cal and mechanical needs include improved sensors and actuators, such as tactile sensors and grippers with closed-Ioop control of grip- ping force. Computational and algorithmic problems in robotics, such as image processing, collision avoidance in a changing en- vironment, and response to fault conditions also need intensive research. Product designs and material and information flows in manufacturing will need major changes to optirn~ze the use of robotic and other flexible automation techniques. Research is needed to establish principles or engineering guidelines that can help determine the optimum degree of flexibility for a manufac- turing system. The man-machine interface must be greatly improved so that people involved in manufacturing can work effectively with vast information flows. Better understanding of human information processing is needed so that the essential information flow be- tween people and machines is speeded up. Interactions must be improved by properly exploiting visual displays in color. Rapid access to large data bases is also needed. For example, production scheduling could be enhanced by computationally quick, dynamic network queueing models, with output at any design point pro- vided via graphic animation. Expert systems techniques may well be helpful in support of interactive decision making; these tech- niques could embody practical knowledge and experience as well as strictly technical data and functions. Such techniques are nec- essary because future flexible manufacturing systems will be too complex for every eventuality to be deterministically programmed. Multiprocessor computer system architectures for real-time control require further research. The data complexity, compu- tational loads, and system reliability requirements of future CIM can be met only with multiprocessor systems. Today, most large real-time computer systems (e.g., those for airline reservations and telephone exchange control) are narrowly optimized to a lirn~ted set of functions. Future CIM systems will have diverse informa- tion flow and real-time response requirements. Computational functions likely will be organized in a hierarchy, and must permit growth to virtually unlimited size, capacity, and reliability. Surely there will be redundancy in both hardware and software. Communication networks for the production environment are evolving rapidly. Simple, reliable, economical techniques are
MANUFACTURING SYSTEMS 225 needed to link diverse systems for design, planning, fabrication, as- sembly, inspection, and testing. The Manufacturing Automation Protocol adopted by General Motors and many other manufactur- ers is a worthwhile step toward standardization in this area. UNIT PROCESSES Research on the individual unit processes that are combined to make up a complete manufacturing sequence has progressed well in many industries. Innovative improvements in materials and unit processes frequently result from the efforts of workers and the engineers who work closely with them. Examples of fields in which more research is needed include processes for forming composite material, net shape forming through precision forging and powder metallurgy, superplastic form- ing and diffusion bonding, direct forming of rapidly solidifying ma- teria~, hot isostatic processing, and evaporative pattern coating. More work is needed for many materials and products to provide an understanding of and models for the relationships among struc- ture, processing, and final properties. In-line instrumentation and control of many unit processes can also be greatly improved. Real-time ctosed-toop process control is becoming more feasi- ble and attractive. The CIM environment makes it possible to adopt processing techniques that are more data-intensive than those commonly used in the past. Thus, in microelectronics man- ufacturing we see the replacement of open-Ioop processes such as wet chemical etching and thermal oxidation with closed-Ioop, electrically controlled processes such as plasma etching, plasma oxidation and anodization, sputtering, and ion milling. Issues that Deternune the Health of Manufacturing Systems Research The previous section described some key system and process technologies that must be a part of CIM. Beyond the forces of technological change are new pressures from corporate manage- ments and from society at large. These new demands stem from increased worldwide competition, higher standards for product quality, concern for environmental effects, and increased demand
226 DIRECTIONS IN ENGINEERING RESEARCH for customized products. Generally, product life cycles are short- ening, leading to shorter planning time, higher ratios of fixed to variable costs, and smaller ratios of selling price to cost. For these reasons, manufacturing has become a larger factor in the strategic health of most businesses and of the national economy as a whole. Manufacturing is changing from the low-risk periphery to the high-risk center stage; from the domain of the unskilled to that of the highly skilled; and from an arena in which skills were learned on the job to one in which advanced formal education is necessary. In the past, many manufacturing processes and products could be understood and evaluated with the unaided human senses; more and more frequently this is impossible. Any assessment of the health of manufacturing research and practice must begin with a recognition of these changes. U.S. manufacturers today are critically short of people who can perform well in the new environment. GOVERNMENT SUPPORT OF MANUFACTURING SYSTEMS RESEARCH Government-supported research on manufacturing systems is only a small fraction of overall public expenditures for engineering research; in turn, engineering research allocations are only a small component of total federal research expenditures. Federal funding under the heading of Programmable Automation" totaled about $82 million in FY84, with more than three-fourths of this money directed toward military programs. The relatively low level of federal research support is one of several serious problems faced by the field of manufacturing systems. Most programs funded by the Department of Defense (DOD) focus on deliverable space and defense technology and as such have relatively short-term orientations. Although these programs are an important too! for cost and quality control in defense systems, they do not and cannot be expected to address the main problems of high-volume commercial manufacturing. Civilian agency programs include those of the National Bu- reau of Standards (NBS), the National Aeronautics and Space Administration (NASA), and the NSF. The research program of the Center for Manufacturing Engineering at the NBS has the aim of developing soundly based standards for integrating het- erogeneous hardware and software in manufacturing systems. It
MANUFACTURING SYSTEMS 227 also aims to develop means and methods for assuring absolute dimensional reliability-in discrete parts production. The work at NBS is of high quality and is directed at two important areas that industry has not effectively addressed. NASA's automation research concentrates on robotic tools for use in space. The research program is small and focused on technologies that are sophisticated by commercial standards. There are occasional spin-offs to commercial manufacturing. The NSF supports a Manufacturing Systems Research Pro- gram, and a number of other programs within the NSF support automation-related research to some degree. NSF-funded research is conducted at universities and covers both unit processes and systems integration. However, these programs are too small to have a major impact. One very positive development is that two of the new NSF Engineering Research Centers will focus on aspects of manufacturing systems. PROBLEMS IN MANUFACTURING ENGINEERING Several difficult problems contribute to the slow progress of industry toward CIM systems. Cause-and-effect relationships are unclear; but certainly the problems are strongly interrelated. A limited professional tradition in manufacturing engineering is at the root of several of the problems we shall enumerate. Many career people in manufacturing lack the scientific and technological knowledge base, as well as the university and professional society contacts outside their own organizations, that are important stim- uli to progress in fields such as aeronautics and electronics. Publi- cations and conferences of the Society of Manufacturing Engineers are strong in their trade components and in traditional elements of manufacturing, but have not been highly elective in promoting research, in stimulating a systems approach to manufacturing, or in forging links between theory and practice. The limited scientific base on which manufacturing engineer- ing is built is a related factor. Other important engineering fields, such as structures, aeronautics, and computers, have generated and then exploited fundamental new knowledge in engineering methods. These fields are guided by growing bodies of scientific principles and engineering techniques that are widely understood, taught, and practiced. The relative absence of such patterns for manufacturing systems reinforces the view that this field is based
228 DIRECTIONS IN ENGINEERING RESEARCH largely on ad hoc empirical methods rather than on a systematic foundation of scientific and engineering principles. To be sure, the empirical knowledge base is important, as it consists of rules dis- covered through long interaction with materials, processes, tools, people, and systems. This "rule-based~ practical knowledge must not be overlooked as fundamental principles are developed and applied. The educational level is traditionally low in manufacturing. The field has little status in the engineering community; new college graduates avoid manufacturing positions. Manufacturing managers frequently are craftsmen promoted up from the shop floor. Formal education has been less necessary for advancement in factory management than in other parts of most companies. Man- agers from such backgrounds must work hard to lead the needed transition to a computer-based systems approach to manufactur- ing. Many more well-educated process specialists and computer engineers will be needed to design and operate the factories of the future. Manufacturing engineers have limited opportunity to build! on the prior work of others. Product development engineers have the opportunity to study the best competing automobiles or comput- ers. Manufacturing engineers have much more difficulty studying their competitors' best factories. Understandably, managements often restrict the flow of information concerning proprietary man- ufacturing methods. As a consequence, manufacturing techniques often vary considerably among firms producing similar end prod- ucts, and there is no arena in which the merits of alternative techniques may be subjected to detailed comparisons. Manufacturing engineering lacks centers of focused research activity akin to the research laboratories of major firms in the com- munications, computer, and chemical industries, and the national centers for research in high-energy physics and materials science. Detailed reporting of the best work via professional conferences and high-quality archival publications has only recently begun. In light of all the abovementioned circumstances, it is easy to under- stand why research on manufacturing is limited in quantity, badly fragmented, and duplicative. Urgently needed generic materials handling and fabrication techniques, flexible manufacturing tools, and information systems standards are appearing much too slowly under the present circumstances.
MANUFACTURING SYSTEMS 229 costs of labor and capital have for many years been higher for U.S. manufacturers than for their strongest international com- petitors. Some well-known attempts at factory automation in the United States failed because of inadequate flexibility in accommo- dating to changes in processes and products. Further, there has been a continuous shortage of engineers qualified to face the chal- lenges implied by computer-integratec] factory automation. These facts have influenced many U.S. manufacturers, particularly in fast-changing fields such as electronics, to choose labor-intensive manufacturing in low-wage areas of the world over automated manufacturing in the United States. Such strategies may well maximize the short-run return on investment. However, they de- flect management's attention from the R&D efforts required to implement modern automated factories for long-term competitive- ness. Implementation of systems such as CIM demands long-term management support. Success in such efforts will require that managers resist the pressure to achieve short-term returns, and that they become personally committed to making the integration process work. Universities have difficulty mounting interdisciplinary efforts needed in manufacturing systems research. Engineering at uni- versities took a strong turn toward quantitative, science-based instruction and research in the 1960s. Research projects have tended to focus on narrow specialty areas in which significant progress might be achieved with the limited resources available to engineering faculty members. Promotions are awarded on the basis of individual achievement; team research (as ~ needed for interdisciplinary fields) has been risky for faculty careers. This problem can be overcome if university administrators devote addi- tional effort to the evaluation of individual contributions to large group efforts. Such differential assessments are routinely made with success in good industrial research laboratories. Credit must be given to all involved when "the whole is greater than the sum of its parts." Manufacturing plants are very expensive to build and very complicated to operate. Small plants cost many millions of dollars and large plants cost many billions. To do meaningful research on the whole manufacturing system requires access to such an operating system, which at this writing can be found only in industry. Thus, university-industry cooperation is vital for the
230 DIRECTIONS IN ENGINEERING RESEARCH conduct of broad-based, interdisciplinary research relevant to the needs of U.S. manufacturing. FORMATION OF RESEARCH CONSORTIA: A HIGH-PRIORITY NEED The United States needs sharply increased R&D activity to speed progress in manufacturing systems engineering. A major objective should be to accelerate the effective application of com- puter technology in order to improve the productivity of U.S. manufacturing and the quality of its output. To achieve this accelerated progress, we believe it is crucial that manufacturers take the initiative in stimulating the formation and strengthening of consortia for research on manufacturing systems. Participants in any such consortium should include numerous in- dustrial firms, and may in addition include universities and federal or state governments. One such consortium is Computer Aided Manufacturing International (CAM-~), based in Arlington, Texas. CAM-T has 150 member companies, but has not been supported at a level adequate to meet all the needs in this field. Examples of strong consortia in other fields are the SRC and the MCC, both of which were formed within the past 3 years. The initiatives for MCC and SRC came from the top management of major corpora- tions in the relevant industries. Actions by Congress and the Jus- tice Department have removed antitrust barriers to the operation of SRC and MCC. SRC supports research and graduate education at universities, with an annual budget of about $20 million. MCC conducts research programs primarily with its own facilities and staff, joined by staff members from participating companies. Its annual budget is about $50 million. The activities of SRC and MCC are complementary, and a number of firms belong to both. Assessment of the Adequacy of New [blent The flow of new engineering talent from the U.S. educational system into manufacturing is inadequate. In reaching this conclu- sion, the Pane! on Manufacturing Systems Research has had access to survey questionnaires; information from universities, technical
MANUFACTURING SYSTEMS 231 societies, and other sources such as the Presidential Young Inves- tigators; and comments from several major corporations. Each input points to shortages of trained American manpower and, depending on the sector, some particular needs. Understandably, because no one sector deals with an overall problem, no one respondent offers overall solutions; all of them quite properly address specific inadequacies with respect to their own needs. The pane! was able to find no positive opinions to offset the prevailing view that there is a shortage of manufacturing · engineering manpower. The pane! believes that the engineering capability of the U.S. manufacturing sector started to slip in the late 1960s. The universi- ties' output of manufacturing engineers was declining and industry continued to fill key positions with personnel Who know machin- ery," such as skilled tradesman. Graduates of 2- and Year insti- tutes of technology provided another attractive option. This led to gaps in position, salary grade, and professional status (obviously coupled to educational level), and eventually to a cultural di~er- ence between degreed engineers who work on product-related items and the few who work in the processes of manufacturing. This gap is real and fundamental to the problem of carrying through total engineering programs, spanning the range from product concept through successful manufacturing. Up-and-down economic cycles and a general focus on short- term corporate financial performance have stifled many attempts at change. There are indications, however, that industrial corpo- rations have now awakened, as evidenced by: . new company-sponsored ways to upgrade engineering per- sonne} in manufacturing by additional training and rewards; . attempts by companies to build manufacturing and engi- neering teams to span the gap between product development and process design; and ~ growing interactions between industry and acadern~a to influence curricula, to foster cooperative work-study arrangements as a part of degree programs, and to support and cooperate in teaching and research. At the same time, universities have been trying to change on their own over the last 12 years, spurred perhaps partly by need, but also by the realization that there is indeed a manu- factur~ng systems discipline. However, there are at present only
232 DIRECTIONS IN ENGINEERING RESEARCH three programs offering degrees in manufacturing engineering that are accredited by the Accreditation Board for Engineering and Technology. Others are hidden under different names or folded into more traditional engineering areas like mechanical engineer- ing. Our feeling is that such nonspecific efforts will not achieve the gains realized in Japan or working in Europe. We applaud these present efforts, but conclude that they should be more specifically supported. Recommendations The pane] believes that a strong, internationally competitive manufacturing component is essential to U.S. domestic and inter- national strength. We also believe that there is now a consensus of opinion in industry, government, and universities favoring sharply stepped-up research and education on manufacturing. Thus, our highest-priority recommendation is that: . Leaders of large U.S. manufacturers should take the initia- tive in stimulating the formation and strengthening of consortia for research on manufacturing systems. Participants in any such consortium should include numerous industrial firms, and may in addition include universities and federal or state governments. We also recommend vigorous actions as follows to improve professional standards in manufacturing engineering: . More specialist workshops, larger and better supported conferences, more refereed professional-level publications, and strong management support for the participation of its engineers in these~activities are needed to place manufacturing engineering more nearly on par with the strongest engineering fields. Academic and industrial leaders must work together through professional so- cieties to achieve these goals. . The newly established program of Engineering Research Centers, fostered by an initiative of the NSF, provides an excel- lent vehicle for university-industry collaboration in research and teaching on manufacturing. Sustained personal and corporate commitments will be essential to the success of these centers.
MANUFACTURING SYSTEMS 233 . The 1984 IBM initiative to upgrade university programs on manufacturing engineering is an example of a highly constructive unilateral corporate initiative. We urge other firms to consider simian initiatives in their fields of interest. The dearth of technical standards for mechanical and elec- trical interfaces and for data transfer has been a severe impediment to progress in manufacturing systems engineering. The General Motors Manufacturing Automation Protocol is a standard now endorsed by many; it must be recognized as an important step forward. For the long term, standards should be established with leadership from professional groups and governments as well as from individual firms. The following actions should be taken by universities and their industrial supporters to improve the state of manufacturing engineering education: . Continued support should be developed and planned for university programs that specialize in coupled manufacturing ed- ucation and research. . Programs shouIc] be defined in which the student can gain a more realistic appreciation of manufacturing applications and the manufacturing systems disciplines. Cooperative work-study programs for engineering students are one good example. . Centers for Manufacturing Excellence should be estate fished to deal with unit processes such as welding, casting, and flex- ible machining, and with computer-integration techniques. These centers could be established via mechanisms that encourage the state governments (financially) to engage local university-industry tearns, such as in the Industrial Technology Institute program in Michigan. . Faculty development programs should be established con- sisting of (reciprocal) personnel exchanges, joint proposals, doc- toral fellowships, grants for research initiation, and Ph.D. thesis projects. A good example of an effort that has evolved over a period of 5 or 6 years into a program like this can be seen at the University of Wisconsin, where an M.S. degree in Manufac- turing Systems Engineering has been established. (This program was stimulated by a grant from IBM.) The University of Wiscon- sin's program is a successful mode! that might well be replicated elsewhere.
234 DIRECTIONS IN ENGINEERING RESEARCH . A professional group should be developed that would estab- lish prestigious awards for contributions in manufacturing, provide matching funds for possible faculty winners of Presidential Young Investigator Awards, and provide seed money awards for course and case study development in manufacturing systems. ~ Industrial support should be encouraged for university re- search and graduate education on manufacturing by establishing additional state and federal government programs, such as Califor- nia's MICRO (Microelectronics and Computer Research Opportu- nities) program. California's program provides for state matching of private grants to universities in support of research agreed to by faculty and counterpart researchers in industry. ~ An effort should be initiated to develop modern, high- quality text material suitable for undergraduate engineering in- struction on manufacturing systems integration. Today such texts are in critically short supply. A parallel situation existed in 1960 with respect to transistor electronics. At that time the Ford Foun- dation and the NSF established and funded the Semiconductor Electronics Education Committee. This group wrote a set of short, high-quality teaching texts that filled this critical need. A similar initiative is needed now to modernize manufacturing engineering curricula. We believe there are enough forward-thinking members in the American university system that an almost endless list of suggestions like the foregoing could be generated. It would make sense to propose such actions only if the academic community could be certain that someone would listen and act on them. Increased support of university program by state and federal governments would certainly be applauded, and might weD be matched, by the industrial sector. Bibliography Air Force Systems Command. Robotic Technology: An Assce~mcnt and Forecast. Wright-Patterson Air Force Base, OH: Air Force Systems Command, 1984. Baum, M. Fact Sheet: Automated Manufacturmg Research Facility (AMRFJ. Washington, DC: National Bureau of Standards, 1983.
MANUFACTURING SYSTEMS 235 Brummett, F. D. The United States Manufacturing Education Experience. Paper presented at the Symposium on Education for the Manufacturing World of the Future, National Academy of Engineering, Washington, DC, September 20, 1984. Charles Stark Draper Laboratory, Inc. Pa,*-on-Dcmand: Manufacturing Tech nology and Tcchr~ology lFansfcr A~c~mcnt Final Report. Report prepared for the Office of Naval Research by the Charles Stark Draper Laboratory, Inc., Cambridge, MA, December 1983. Haller, H. D. Examples of Uruvcraity-Indu~try-Goucrnmcnt Collaborations. Ithaca, NY: Office of Vice President for Research and Advanced Studies, Cornell University, August 1984. Lardner, J. F. Industry and Interdisciplinary Teams: Experience and Expec- tations. In Information and Technology Exchange Among Engineering Research Ccntere and Industry. Report of a workshop held by the Cross-Disciplinary Engineering Research Committee, National Research Council, Washing- ton, DC, 1985. McNinch, S., Jr. Engincenr~g An Expar~ded and More Active Role for NSF. Washington, DC: National Science Foundation, 1985. Meade, W. P. Lee National Bureau of Standard,' Automated Manufacturing Rc- ~earch Facility (AMRF}An Analysis of Its Impact. Chapel Hill, NC: Man- agement Collaborative Group, November 1984. National Academy of Engineering. U.S. Leadership in Manufacturing. Pro- ceedings of a Symposium at the Eighteenth Annual Meeting, National Academy of Engineering, Washington, DC, November 4, 1982. National Research Council. The U.S. Machine Tool Industry and the Defense Industrial Base. Committee on the Machine Tool Industry, Manufacturing Studies Board, National Research Council. Washington, DC: National Academy Press, 1983. National Research Council. Computer Integration of Engineering Design and Production: A National Opportunity. Washington, DC: National Academy Press, 1984. National Science Foundation. Workshop on Materials Processing. Report of a Workshop Sponsored by the National Science Foundation and orga- nized by the Processing Research Institute, Carnegie-Mellon University, Pittsburgh, PA, October 3~31, 1975. National Science Foundation. Workshop on Materials Processing. Report of a Workshop sponsored by the National Science Foundation and orga- nized by the Processing Research Institute, Carnegie-Mellon University, Pittsburgh, PA, June 1976. Office of Technology Assessment. U.S. Industrial Competitiveness A Comparison of Steel, Electronics and Automobiles. Washington, DC: U.S. Congress, Office of Technology Assessment, July 1981. Office of Technology Assessment. Computerized Manufacturing Automation: Employment, Education and the Workplace (OTA-CIT-2353. Washington, DC: U.S. Congress, Office of Technology Assessment, April 1984. Procecclinge of the Twelfth Annual lki-Scrvicc Manufacturing Technology Confcrcnec. Bal Harbour, FL, October 19-23, 1980.
236 DIRECTIONS IN ENGINEERING RESEARCH Science Applications, Inc. Parts on Demand- Evaluation of Approaches to ~4chieuc Flexible Manufacturing Sy~tcrru for Navy Pa,* on Demand Vol. I. Prepared for the Naval Supply Systems Command and Office of Naval Research by Science Applications, Inc., Robotic and Automation Division, McLean, VA, February 1984. Semiconductor Electronics Education Committee (seven-volume set). New York, NY: Wiley, 1964-1966. Simpson, J. A., R. J. Hocken, and J. S. Albus. The Automated Manufac- turing Research Facility of the National Bureau of Standards. Journal of Manufacturing Systems 1~1~:17-32, 1982. Society of Manufacturing Engineers. Final Report on The Manufacturing Engineer: Past, Pranced and F~turcn to Society of Manufacturing Engineers. Prepared by Battelle Columbus Laboratories, Dearborn, MI, 1979. Society of Manufacturing Engineers. Directory of Manufacturing Research Needed by Indwiry. Dearborn, MI: Society of Manufacturing Engineers, 1982. Society of Manufacturing Engineers. Directory of Manufacturing Education Programs in Collegc* Univcr~itice, and Technical Institute: Engineering, Engineering Tcchnology, Indwirial Tcchnology, 19~-1985. Dearborn, MI: Education Department, Society of Manufacturing Engineers, 1984.
MANUFACTURING SYSTEMS Appendix Responses to the Engineering Research Board's Request for Assistance Mom Universities, Professional Societies, and Federal Agencies and Laboratories 237 Requests for assistance were sent by the Engineering Research Board to a number of universities, recipients of Presidential Young Investigator Awards, professional societies, and federal agencies and laboratories in order to obtain a broader view of engineering research opportunities, research policy needs, and the health of the research community. Some of the responses included com- ments on engineering research aspects of manufacturing; these were reviewed by this panel to aid in its decision-making process. The pane] found the responses to be most helpful and wishes that it were possible to individually thank all those who took the time to make their views known. Because that is not practical, we hope nevertheless that this small acknowledgment might convey our gratitude. Responses on aspects of manufacturing were received from in- dividuals representing 45 different organizations, listed in Table A: 26 universities (including 9 represented by recipients of NSF Presi- dential Young Investigator Awards), 10 professional organizations, and 9 federal agencies or laboratories. Some comments covered specific aspects of the panel's scope of activities, whereas others provided input on a variety of subjects.
238 DIRECTIONS IN ENGINEERING RESEARCH TABLE A-1 Organizations Responding to Information Requests Relevant to Manufacturing Systems Research UNIVERSITIES Brigham Young University Carnegie-Mellon University Clarkson University Drexel University Duke University Lehigh University Massachusetts Institute of Technology North Carolina State University Northwestern University Princeton University Purdue University Rensselaer Polytechnic Institute Texas A&M University University of Connecticut University of California, D avis University of California, Los Angeles University of IllinoisUrbana/ Champaign University of Iowa University of Kansas University of Maryland University of Michigan University of Minnesota University of Oklahoma University of Pennsylvania University of Rochester University of Utah PROFESSIONAL ORGANIZATIONS American Institute of Aeronautics and Astronautics American Institute of Chemical Engineers American Society of Civil Engineers American Society of Mechanical Engineers The Institute of Electrical and Electronics Engineers, Inc. Institute of Industrial Engineers Industrial Research Institute Society of Engineering Science, Inc. Society of Manufacturing Engineers Society of Naval Architects and Marine Engineers AGENCIES AND LABORATORIES Air Force Institute of Technology Air Force Office of Scientific Research Army Materials and Mechanical Research Center Army Research Office Lawrence Livermore National Laboratory NASA Goddard Space Flight Center NASA Langley Research Center Oak Ridge National Laboratory Sandia National Laboratories
