Publications

We have reviewed ten major public problems challenging our Nation as it enters the new millennium. These are defense, healthcare costs, education, aging population, energy and environment, crime, low productivity growth services, income distribution, regulations, and infrastructure. These problems share several features. First, each is so large, if it were soIved; it would have major impact on the U.S. economy. Second, each is resident in a socioeconomic system containing non-linear feedback loops and an adaptive human element. Third, each can only be solved by our political system, yet these problems are not responsive to piecemeal problem solving, the approach traditionally used by policy makers. However, unless each problem is addressed in the context of the system in which it resides, the solution maybe worse than the problem. Our political system is immersed in reams of disconnected, unintelligible information skewed by various special interests to suggest policies favoring their particular needs. Help is needed, if rational solutions that serve public interests are to be forged for these ten probIems, The simulation and modeIing tools of physical scientists, engineers, economists, social scientists, public policy experts, and others, bolstered by the recent explosive growth in massively parallel computing power, must be blended together to synthesize models of the complex systems in which these problems are resident. These models must simulate the seemingly chaotic human element inherent in these systems and support policymakers in making informed decKlons about the future. We propose altering the policy development process by incorporating more modeling, simulation and analysis to bring about a revolution in policy making that takes advantage of the revolution in engineering emerging from simulation and modeling. While we recommend major research efforts to address each of these problems, we also observe these to be very complex, highly interdependent, multi-disciplinary problems; it will challenge the U.S. community of individual investigator researchers to make the cultural transformation necessary to address these problems in a team environment. Furthermore, models that simulate future behavior of these complex systems will not be exacq therefore, researchers must be prepared to use the modeling and simulation tools they develop to propose experiments to Congress. We recommend that ten laboratories owned by the American public be selected in an interagency competition to each manage and host a $1 billion/yertr National effort, each focused on one of these ten problems. Much of the supporting research and subsystem modeling work will be conducted at U.S. universities and at private firms with relevant expertise. Success of the Manhattan Project at the middle of the 20th century provides evidence this leadership model works.

Federal laboratories have successfully filled many roles for the public; however, as the 21st Century nears it is time to rethink and reevaluate how Federal laboratories can better support the public and identify new roles for this class of publicly-owned institutions. The productivity of the Federal laboratory system can be increased by making use of public outcome metrics, by benchmarking laboratories, by deploying innovative new governance models, by partnerships of Federal laboratories with universities and companies, and by accelerating the transition of federal laboratories and the agencies that own them into learning organizations. The authors must learn how government-owned laboratories in other countries serve their public. Taiwan`s government laboratory, Industrial Technology Research Institute, has been particularly successful in promoting economic growth. It is time to stop operating Federal laboratories as monopoly institutions; therefore, competition between Federal laboratories must be promoted. Additionally, Federal laboratories capable of addressing emerging 21st century public problems must be identified and given the challenge of serving the public in innovative new ways. Increased investment in case studies of particular programs at Federal laboratories and research on the public utility of a system of Federal laboratories could lead to increased productivity of laboratories. Elimination of risk-averse Federal laboratory and agency bureaucracies would also have dramatic impact on the productivity of the Federal laboratory system. Appropriately used, the US Federal laboratory system offers the US an innovative advantage over other nations.

SEMATECH was established in 1987 for defense and economic reasons to help the U.S. regain a competitive posture in semiconductor manufacturing. For 10 years SEMATECH was jointly funded by the federal government and semiconductor manufacturing companies representing 85 percent of the U.S. semiconductor industry. SEMATECH has spent about 80 percent of these funds on activities intended to produce useful results between 1 and 3 years. Very early in the establishment of SEMATECH, its members determined that their first priority would be to strengthen their U.S. based suppliers of semiconductor manufacturing equipment. This has been the primary thrust of SEMATECH. SEMATECH first held some 30 workshops on a broad set of technical topics to assess the needs and opportunities to help the industry recover. These workshops scoped manufacturing areas where SEMATECH should focus. These early meetings were an early form of what later came to be termed roadmapping. The scope of R&D needs identified in these workshops well exceeded what SEMATECH could hope to accomplish with its $200 million annual budget. Wayne Johnson of Sandia participated in five of these workshops and used the knowledge gained as the basis for proposals later submitted to SEMATECH on behalf of Sandia. In the fall of 1989 the SETEC program was established at Sandia to support SEMATECH. This was initially a funds-in, work-for-others project that was fully funded by SEMATECH. Thus, the early work was entirely focused on SEMATECH`s needs. Later in the program when SEMATECH funds were supplemented by Department of Energy Cooperative Research and Development funds, attention was given to how this project would benefit Sandia`s defense microelectronics program.

This paper presents an analysis of how the US government responded to the concern in the 1980`s that US companies were experiencing problems of competitiveness in international markets. By the mid 1980`s there was great and growing concern throughout the US that US companies were experiencing difficulties in international competition. Pressure on Congress to take action came from constituents seeking jobs and companies that would directly benefit (this usually means receive public money) from programs that Congress might initiate. The fact that most constituent calls to Congress were about job creation was lost in the on-rush of R&D performers seeking funds for their favorite R&D project. In response, Congress created the Advanced Technology Program, the Technology Transfer Initiative, and the Technology Reinvestment Project, expanded the responsibilities of ARPA/DARPA, increased funding for the Small Business Initiative, expanded the Manufacturing Extension Partnership, funded SEMATECH, and increased NSF funding for basic research at universities. Many of these programs were later criticized for being industrial welfare and several were cut-back or stopped. Retrospective analysis shows that few of these programs addressed the root cause of competitiveness difficulties. In fact, by the time most of these programs were in place, US companies were well on their way to correcting their competitiveness problems. In addition, few were relevant to companies` often expressed concerns about workforce training, regulatory costs, and access to foreign markets. Twenty percent reductions in health care costs, regulatory costs, and education costs could annually pump $500 billion into the US economy and make companies operating in the US much more competitive in international markets.

This report addresses the need for government sponsored R and D to address real public problems. The motivation is that a public benefit of the money spent must be demonstrated. The areas identified as not having appropriate attention resulting in unmet public needs include healthcare cost, cost and benefits of regulations, infrastructure problems, defense spending misaligned with foreign policy objectives, the crime problem, energy impact on the environment, the education problem, low productivity growth industry sectors, the income distribution problem, the aging problem, the propagation of disease and policy changes needed to address the solution of these problems.

Federal R and D must be principally focused on solving public problems that the marketplace is failing to address. With few exceptions programs must be supported by roadmaps that show how the R and D is linked to public outcomes. Federal R and D and those who perform it must be judged in terms of the public outcomes. The overarching issues of federal R and D policy, what it should address, how to manage it, who should perform it, how to perform it, what works best, etc. are highly complex and lack a strong theoretical foundation. (In fact, the linear, assembly-line model used by policymakers is wrong.) It is time that policymakers recognize and acknowledge the uncertainty of their work and conduct a wide array of policy experiments (the authors consider SEMATECH such an experiment) that are supported by public outcome metrics. In addition to making federal R and D better address public needs, such an approach to policy making could raise the public`s interest in T and S policy. Of course, as in any experiment the results may be measured and if failures aren`t observed, it is likely that policies lack vision and imagination. It is time to abandon the budget driven federal R and D system where performers of federal R and D are treated as constituents, and replace it with a public problem--public outcome driven system where public problems are prioritized and the budget is distributed to agencies according to these priorities.

A review of US and Japanese experiences with using microelectronics consortia as a tool for strengthening their respective industries reveals major differences. Japan has established catch-up consortia with focused goals. These consortia have a finite life targeted from the beginning, and emphasis is on work that supports or leads to product and process-improvement-driven commercialization. Japan`s government has played a key role in facilitating the development of consortia and has used consortia promote domestic competition. US consortia, on the other hand, have often emphasized long-range research with considerably less focus than those in Japan. The US consortia have searched for and often made revolutionary technology advancements. However, technology transfer to their members has been difficult. Only SEMATECH has assisted its members with continuous improvements, compressing product cycles, establishing relationships, and strengthening core competencies. The US government has not been a catalyst nor provided leadership in consortia creation and operation. We propose that in order to regain world leadership in areas where US companies lag foreign competition, the US should create industry-wide, horizontal-vertical, catch-up consortia or continue existing consortia in the six areas where the US lags behind Japan -- optoelectronics, displays, memories, materials, packaging, and manufacturing equipment. In addition, we recommend that consortia be established for special government microelectronics and microelectronics research integration and application. We advocate that these consortia be managed by an industry-led Microelectronics Alliance, whose establishment would be coordinated by the Department of Commerce. We further recommend that the Semiconductor Research Corporation, the National Science Foundation Engineering Research Centers, and relevant elements of other federal programs be integrated into this consortia complex.

The core problem with the US health care system is -- it already costs to much and the rate of its cost growth is cause for further alarm. To deal with these, regulators must introduce incentives for health care providers to reduce costs and introduce incentives that make consumers of health care services concerned about the costs of the services they demand. Achievement of these regulatory goals will create opportunities for the introduction of innovations, including revolutionary new technology, that can lead to major reductions in costs. Modeling of health care system inputs, outputs, transactions, and the relationships between these parameters will expedite the development of an effective regulatory process. This model must include all of those major factors that affect the demand for health care and it must facilitate benchmarking health care subsystems against the most efficient international practices.

Heavy particle radiation can produce upsets in digital circuits as well as trigger burn out or breakdown in power MOSFETs and MNOS nonvolatile memories. Latch-up may also be stimulated by heavy ions. This report covers work done on the effects of heavy particle radiation on PN junctions, CMOS inverters, CMOS latch, MOSFET and non-volatile memories. 15 refs., 3 figs.