
Key Issues This page: Plemmons (JPBM) More Topics Key Questions Problems & Opportunities Equity & Access Minorities Women Disabilities NonEnglish Job Market Public Policy Typesetting Public Understanding AMS Ethical Guidelines 
What Does the Government Get For Its Investment in Basic Research?Joint Policy Board for MathematicsTestimony on FY 1997 Appropriations for the Department of Defense
This testimony was delivered by Prof. Robert Plemmons on 1 May 96 before the House Appropriations subcommittee on National Security and on 18 June 96 before the Senate Appropriations subcommittee on Defense. For more information on JPBM's policy and government affairs activities, please contact Lisa Thompson, JPBM Congressional Liaison. For more information on the work described by Prof. Plemmons, visit his home page. A summary of this testimony is to be found in the June 1996 SIAM NEWS, p.2. Good afternoon, Mr. Chairman and Members of the Subcommittee. I am Robert Plemmons, Z. Smith Reynolds Professor of Mathematics and Computer Science at Wake Forest University. I appreciate this opportunity to comment on FY 1997 appropriations for the Department of Defense. I speak on behalf of the Joint Policy Board for Mathematics, which represents three associations of mathematical scientists whose concerns encompass fundamental and interdisciplinary research; the applications of the mathematical sciences in science, engineering, and industry; and mathematics education at all levels. Today I would like to address DOD's investment in basic research in support of the defense mission, an activity funded out of the 6.1 R&D account. Mr. Chairman, JPBM urges Congress to continue recognizing DOD's investment in basic research, including research conducted at universities, as an integral and foundational component of R&D efforts undertaken to meet the Nation's defense needs. We ask that you provide full funding for this investment, approximately $1.16 billion, including about $820 million for Defense Research Sciences, in FY 1997. Since World War II, the United States has relied on the superiority of its military technologies to achieve its national security objectives. Basic research is essential to the process of developing new technologies, improving existing ones, employing them as effectively as possible, and therefore maintaining superiority over the long term. By engaging the Nation's research universities in this process, DOD has access to firstrate researchers and the latest discoveries in pursuit of its R&D objectives. The origins of many key defense technologies can be traced to DOD support for basic research conducted at U.S. academic institutions. The defense agencies that sponsor basic researchoethe Army Research Office, the Office of Naval Research, the Air Force Office of Scientific Research, and the Defense Advanced Research Projects Agencyoehave an excellent track record for making decisions about which areas are vital to DOD's technology goals and which researchers are best able to mine the scientific opportunities for contributions to national security. Rigorous internal decisionmaking processes guide these investments to ensure both scientific excellence and consistency with DOD's strategic priorities. They take full advantage of the robust U.S. research system, supporting work at a mix of universities, governmental laboratories, and sometimes in cooperation with industry. These thoughtfully planned investments need stable yeartoyear funding to avoid curtailing the scope of promising research that DOD has identified as relevant to its mission. Let me demonstrate the importance of basic research to DOD's mission with some examples from my own field, the mathematical sciences. DOD's support for research in the mathematical sciences is most often motivated by the need to enhance the effectiveness, efficiency, and reliability of US defense hardware and capabilities, for example, communications systems, cryptographic capabilities, and smart weapons technology. The field also underpins many critical and emerging technologies that are crucial to national security, like materials, robotics, and information technologies. In many of the mathematical areas relevant to defense, university researchers are in an excellent position to make very productive advances. DODsupported mathematical researchers often work very closely with their counterparts in DOD laboratories. I would also like to stress that universitybased defense research makes an important contribution to mathematics, science, and engineering education, ensuring that we have a cadre of graduates who are acquainted with defenserelated research. I have been working with DOD support since 1973. In recent years my work has centered around the improvement of groundbased, airtoair, and airtoground imaging systems through the development of adaptive optics and image postprocessing techniques. The problem is that atmospheric turbulence and other optical aberrations impose serious limitations on the quality of data captured by optical imaging systems, whether it's satellites taking pictures of the earth or groundbased telescopes taking pictures of the sky. Adaptive optics techniques involve measuring the atmospheric distortions and making realtime adjustments in how the images are captured, resulting in data that is of significantly higher resolution than it otherwise would be. For instance, telescopes can be built with deformable mirrors that change their shape depending on the level and nature of the atmospheric effects they need to overcome. Similar measurements and adjustments are used by imaging hardware on observational satellites. The accuracy of these images can be restored further, subsequent to their capture, by employing even more sophisticated computations. The resulting improvement in image clarity is easily demonstrated in "before" and "after" photographs. Each stage of this "deblurring" processoemeasuring the distortion, adjusting the equipment, and restoring the imagesoerelies on rigorous mathematical and computational techniques to determine what compensations are needed to generate images of the best possible clarity. These techniques are being continually refined through frontier mathematical research, including my own, supported by the Air Force Office of Scientific Research. Another area in which mathematical research is leading to very useful results is the handling and analysis of large volumes of data, something defense personnel and technology are faced with every day. For instance, let's again focus on observational satellites taking pictures of the Earth's surface. Those images have to be conveyed somehow back down to Earth before we can use them. The images are compressed into digital form and a digital signal is beamed to a receiver. It is imperative that the data be compressed in a way that allows the receiver to decompress it into usable form that accurately reflects the original data. In the mid1980s mathematical scientists invented a mathematical method to represent and convey signal and image data in a very efficient way. It enables compressed data to be retrieved and used as needed, very quickly, at different scales of resolution, and with little loss of accuracy, so that the most precise data can be used in realtime situations. DOD scientific personnel recognized early that this was a breakthrough that would lead to significant improvements in defense operations. The Air Force and DARPA continue to support research into these mathematical techniques, which go by the name of "wavelets", and DOD is putting them to work in a variety of ways. For instance, it appears that one day the techniques can be used for automatic detection and recognition of targets, which is simply not possible with older methods. That is, the analysis of data while it is still compressed can be automated, so that when, say, a satellite finds something that might be of interest to DOD, the appropriate personnel can be instantly alerted to its existence. Presently, separate and cumbersome processes must be used to analyze collected data after it is decompressed. Incidentally, it turns out that the FBI found wavelets to be the ideal method for compressing fingerprint images, because the fine details of the images are preserved so well. Here is another instance in which powerful mathematical techniques, used in combination with advanced computers, are providing dramatic solutions to key national security problems. DOD needs to be able to predict the scattering of radar signals from aircraft so their vulnerability to detection can be analyzed and so that new aircraft with low radar visibility can be designed. The programs currently used to make the predictions take a long time to run, so they are only practical when applied to small components of an aircraft, rather than to the entire aircraft. Mathematical researchers are working to develop computational codes that can predict the radarscattering patterns of large objects more efficiently. The new, faster techniques will make it possible to compute patterns across major components or entire aircraft in an acceptable period of time. For example, as the size of the object to be analyzed increases by a factor of 10, the computing time using existing methods increases by a factor of 1000, while the newer algorithms would run in just one percent of that time. I hope these examples demonstrate how research in the mathematical sciences leads to time and moneysaving technologies for use in critical defense systems. DOD's ongoing pursuit of new developments in basic science will no doubt continue to yield these kinds of results. So let me again urge you to continue your support for DOD's investment in basic research and for the contributions of universitybased researchers. Thank you for this opportunity to express our views for the record regarding FY 1997 appropriations. I would be pleased to answer any questions you might have.

[Privacy Policy] [Terms of Use]
Home  The Math Library  Quick Reference  Search  Help