CISE - Advisory Committee
DIRECTORATE FOR COMPUTER AND INFORMATION SCIENCE AND ENGINEERING ADVISORY COMMITTEE (CISE AC)
April 4-5, 2000 Radisson Barcelo Hotel Washington, D.C.
Dr. Forest Baskett, Chair, New Enterprise Associates, Menlo Park, CA
Dr. James Allen, Dept. of Computer Science, University of Rochester, Rochester, NY
Dr. Charles Bennett, T. J. Watson Research Center, Yorktown Heights, NY
Dr. Christine Borgman, Information Studies, University of California, Los Angeles, CA
Dr. David DeWitt, Computer Science Department, University of Wisconsin, Madison, WI
Dr. James Flanagan, Research, Rutgers University, Piscataway, NJ
Dr. Louis Gomez, School of Education and Social Policy, Northwestern University, Evanston, IL
Dr. Mary Jane Irwin, Dept. of Computer Sciences & Engineering, Penn State University, University Park, PA
Dr. Leah Jamieson, School of Electrical & Computer Engineering, Purdue University, West Lafayette, IN
Dr. Sid Karin, University of California-San Diego, La Jolla, CA
Dr. Edward Lazowska, Dept. of Computer Science & Engineering, University of Washington, Seattle, WA
Dr. Cherri Pancake, Department of Computer Science, Oregon State University, Corvallis, OR
Dr. Christos Papadimitriou, Computer Science Division, University of California-Berkeley, Berkeley, CA
Dr. Daniel Reed, Department of Computer Science, University of Illinois, Urbana, IL
Dr. Herbert Schorr, University of Southern California Information Sciences Institute, Marina del Rey, CA
Dr. Jonathan Smith, University of Pennsylvania, Philadelphia, PA
Professor Richard N. Taylor, University of California-Irvine, Irvine, CA
Dr. Andries van Dam, Brown University, Providence, RI
Dr. Paul Woodward, Astronomy Department, University of Minnesota, Minneapolis, MN
Dr. Paul N. Hale, Jr., Louisiana Tech University, Ruston, LA (Ex Officio Member, CISE Advisory Committee)
Dr. David Messerschmitt, University of California-Berkeley, Berkeley, CA
Dr. Richard Rashid, Advanced Technical Research, Microsoft, Redmond, WA
Dr. Eugene H. Spafford, Computer Sciences, Purdue University, West Lafayette, IN
A meeting of the Directorate for Computer and Information Science and Engineering’s Advisory Committee (CISE AC) was held at the Radisson Barcelo Hotel in Washington D.C. on April 4-5, 2000.
April 4, 2000
WELCOME AND INTRODUCTIONS
Dr. Forest Baskett, Chair, called the meeting to order at 8:30 a.m. and welcomed the participants. He introduced three new members to the CISE AC: Dr. Charles Bennett, Dr. Jonathan Smith, and Professor Richard N. Taylor. He announced that a previous member of the CISE/AC, Dr. John Hennessy, had been named President of Stanford University. He introduced the minutes from the October 1999 CISE AC and invited comments. A motion to accept the minutes was made and seconded. He then introduced Dr. George Strawn, Executive Officer, CISE, who provided an overview of the agenda which would include all four NSF initiatives: Information Technology Research (ITR), Nanoscale Science and Engineering, the 20th Century Workforce, and Biocomplexity. Dr. Strawn noted that all of the presentations would be available on the web. He also referenced a list of CISE vacancies and asked the CISE AC for recommendations.
UPDATE OF INFORMATION TECHNOLOGY RESEARCH
Dr. Michael Lesk, Division Director, IIS, reported on the CISE ITR budget and said that Congress had appropriated $90 M for fundamental research and $36 M for terascale computing in FY2000. An additional $100 M is requested for FY2001. The current solicitation highlights major topic areas for research: software; human-computer interface; information management; advanced computational science; scalable information infrastructure; social and economic implications; and revolutionary computing. Submissions included 920 preproposals for larger awards and 1,100 proposals for smaller awards. The focus was on fundamental research on IT. Proposals were received from all parts of the US with the top ten universities submitting 257 proposals. Only 68 institutions were encouraged to submit full proposals.
Dr Lesk itemized the goals considered in awarding grants and asked how CISE could encourage really innovative work since proposers complain that if they submit work that is too innovative, the review panel rejects it. PITAC was explicit on where money should be spent–on the best proposals. Items to consider are:
- The proposals must balance educational impact and research.
- Extending the community.
- In regard to elitism v. populism, PITAC suggested a need for more top universities. Others suggest encouraging more submissions from less well-known universities.
- Cost-sharing is a requirement.
- Importance of multidisciplinary efforts and multi-institutional work.
- Collaboration with industry, natinal labs and international collaborations.
- Balance between large and small awards.
The CISE AC discussed how difficult it would be to satisfy these issues. Concern was expressed about the lack of innovation the group discussed ways to get more innovative proposals such as adding something to the review criteria for ranking the risk.
PARTNERSHIPS FOR INNOVATION
Dr. Robert Borchers, Division Director, ACIR, presented an overview of a new program, Partnerships for Innovation. He served on the working group which established it in response to Congress’s request. It promotes and supports increased collaborations among universities and between universities and Federal, state, and regional organizations and the private sector. A goal is to focus on institutions and states which have the least amount of Federal funding and to drive innovation. The Partnerships for Innovation program provides outreach to non-research intensive universities that have limited participation in research and education grant programs. The solicitation focuses on NSF’s strategic goals:
- Ideas - Discovery across frontiers connected to learning, innovation and service to society;
- People - A diverse, internationally competitive and globally-engaged workforce;
- Tools - Accessible, state-of-the-art information bases and shared tools.
The solicitation invites a diverse set of awards and submission is limited to two per institution. NSF will make 15-25 awards for $300-$600 K for 2-3 years. The deadline for submission is July 6, 2000.
INTEGRATIVE GRADUATE EDUCATION AND RESEARCH TRAINEESHIPS (IGERT)
Lawrence Goldberg and Wyn Jennings, co-chairs of the administering working group, presented an overview of the IGERT program, which has been in existence since 1997. It is Foundation-wide and addresses the strategic goals of the NSF. It supports graduate students through traineeships and focuses on multidisciplinary research that is problem-oriented, not discipline centered; innovative approaches to education and training across disciplines; and preparing students with skills for the 21st century. It addresses cutting-edge research in science and engineering and innovative graduate education; fosters professional and personal skills and a global perspective; provides opportunities for both academic and non-academic careers; and links graduate research with industry, national laboratory, international, and other settings. It provides a strategy for recruiting and mentoring under-represented groups in order to develop a diverse science and engineering (S & E) workforce.
Some changes have been made for the FY2000 solicitation but there is still a requirement that all stipend recipients be citizens or permanent US residents. A CISE AC member commented that because of this requirement, good foreign students are not able to get funding and this has resulted in a drop in the number of foreign students getting Ph.Ds. Dr. Goldberg said that if we want our US students to be strong in technical areas, we have to start in K-12 and then mentor undergraduate programs to encourage them. Members of the CISE AC agreed that minority groups should be encouraged because 50 percent of minority students drop out of Ph.D. programs after they take their qualifying exams. The AC asked if there is funding for long-term tracking of success of the students in the IGERT program. Dr. Jennings said that there would be web-based data collection in place and every student will be followed for two years following completion of the program.
The program has been very successful as indicated by its recognition in the university community. There have been 57 awards since its inception, and funding has increased from $30 M per year to $54 M by FY2004. CISE participation has increased since 1998, but the CISE AC should be proactive to gain even more CISE involvement. Dr. Adrion noted that Division of Experimental and Integrated Activities is the contact division for IGERT.
NANOSCALE SCIENCE AND ENGINEERING INITIATIVE
Dr. Mihail Roco, ENG, made a presentation about the Nanoscale Science and EngineeringInitiative, which started informally in 1996. In 1998, the National Science and Technology Council (NSTC) established the Interagency Working Group on Nanoscience, Engineering, and Technology (IWGN), which comprises eight Federal agencies. The FY1999 Federal budget for nanotechnology research was $260 M. The IWGN recommended doubling it for FY2001. The President’s Committee of Advisors on Science and Technology (PCAST) Nanotechnology Panel identified priorities for the initiative.
Dr. Roco said that nanotechnology has given us the ability to manipulate matter economically and provided new tools which allow for fundamental discoveries. Miniaturization is of interest because it is faster, requires less material, and less energy. More important are novel properties, new phenomena, and its ability to create new molecules and structure. The fields of physics, chemistry, engineering, and biology all use the same tools to work in nanoscale.
Long-term, high-risk research is needed and there has been a high level of enthusiasm in the scientific community. Several discoveries have already occurred, such as nanolayers in hard disk heads, nanostructured chemical catalysts, advanced drug delivery systems, nanolayers with selective optical barriers, hard coatings, dispersions, chemical and biodetectors, new generations of lasers, and systems on a chip. Emerging nanoelectronics includes extending silicon technology, new devices, single electron tunneling, resonant tunneling diodes, etc. Dr. Roco said that although the US does not have a commanding lead in nanotechnology, we do have good infrastructure and are moving fast in the area.
R & D funding for the nanotechnology initiative is for $217 M for 2001. Dr. Roco stated that three of the focus areas of the initiative address CISE with a specific role under Device and System Architecture. Dr. Roco said that electronics will probably take over but for now the research is more involved with chemistry, physics, and biology. He said the program announcement would be out in August and he asked for input from the CISE community. A question was raised about the socioeconomic effect on the workforce and also the use of technology transfer to accelerate the research, suggesting that the social study of technology should move it from research to application. Dr. Roco said that industry interaction is planned. Dr. Baskett asked that Dr. Roco come back and talk at the next meeting.
WORKING LUNCH: CISE AC ON "WHAT'S NEW AND EXCITING?"
CISE AC members were asked to share what is new and exciting about their research during a working lunch.
- Christine Borgman, University of California, talked about digital libraries, information retrieval and human/computer interaction, and the study of network relationships among documents on line. Publishers are forming consortiums and working on technical issues of crossing links between publications and digital libraries.
- Charles Bennett, T.J. Watson Research Center, is working on quantum information theory, using quantum computing to create physical devices that will allow problems to be solved faster and provide more information storage. It represents a redrawing of the line between physics and computer science.
- Andries van Dam, Brown University, is trying to get a consortium together to fund a $5 M project in educational technology. He is concerned that the educational software being produced will not work on the next generation computers. He is particularly interested in simulation-based software and gave two examples: virtual laboratory (including lectures) and cognitive tutors.
- Christos Papadimitriou, University of California, talked about improvements in the theoretical field and the challenges of quantum computing. He also said that California is providing $100 M each to fund three large labs for science and technology.
- Edward Lazowska, University of Washington, has been working on networking engineering for the past six months, trying to improve the communication in the Northwest, by either extending the vBNS or getting DARPA’s NGI testbed into Seattle. He has done demos of gigabit per second networking from desktop to desktop.
- James Allen, University of Rochester, is working on human/computer interaction using natural language based systems. He said that the most exciting thing that has happened is the hardware is now fast enough so that speech recognition in real time is almost as fast as human conversation.
- James Flanagan, Rutgers University, is working on a protein macromolecule data base where chemists and biologists deposit data using surface physics which relates to microscopic data storage at the molecular level, allowing for the possibility of unlimited data storage. He’s also interested in human-computer interaction and described some examples in telemedicine (for telerehabilitation and telediagnosisis).
- Jonathan Smith, University of Pennsylvania, is doing experimental work in networking and security, (introducing code into networking structure) and addressing the issue of how to detect the presence of sensors.
- Paul Woodward, University of Minnesota, said his main interest is in getting more and more commodity in visualization driven by SGI . He talked about connecting hundreds of video boards used in computer games. He’s also interested in harnessing a large number of CPUs to increase memory.
- Sid Karin, University of California, talked about their new IBM processing mode system which includes 1,200 processors. He said getting them all to work is a problem, not only because of the hardware, but the software as well. It was used in the rendering of a flight for the Orion nebula at the Hayden Planetarium.
- Herbert Schorr, University of Southern California, is involved in the consortium for the gigabit on a desktop; working on a project on the security issues of ubiquitous sensors and an Army project to study how Hollywood relates reality, which will be used in training Army personnel.
- Dan Reed, University of Illinois, is working on how to derive distributed applications across a wide area of computational grids. He’s also interested in attracting graduate students into computer science and suggests concentrating on short-term commitments rather than long-term.
- Cherri Pancake, Oregon State University, is working on how to structure software so that the model is more intuitive, constructing models to illuminate the learning curves, and how to transfer a sense of reality to a remote user.
- Leah Jamieson, Purdue University, is working on the integration of speech and language recognition and also integrating speech and gesture, to have different modalities of access such as developing devices to point for people who can’t talk.
21st CENTURY WORKFORCE INITIATIVE
Dr. Judith Sunley, Assistant Director, EHR, listed areas of emphasis in EHR for FY2001, which are: research on the science of learning; enhancing Science, Mathematics, Engineering, and Technology (SMET); near-term workforce needs; broadening participation in SMET; and networking and access (digital libraries). The FY2001 budget will fund: Research on Learning and Education; Centers for Learning and Teaching; Graduate Teaching Fellows in K-12; Advanced Technological Education; Scholarships for Service; Tribal College Initiative; and a National SMET Digital Library. She said that there is great interest in the functional range of cognitive development and learning processes that will help teachers. She addressed the relevance of CISE to the areas of emphasis, the relative roles and responsibilities of the CISE and EHR Directorates, and the distinctions between the IT Initiative and the 21st Century Workforce Initiative.
Dr. Sunley said that NSF has tried to enhance connections between K-12 and higher education. Also addressed was what is NSF’s role, what interplay should there be between the states and the Federal government and how creative approaches can be developed. In order to define the role of CISE and to get some creative thinking from the math and science community, members of the CISE AC will meet with EHR AC. CISE will provide some guidance on developing a handbook for evaluation of the program. A CISE AC member commented that Dr. Sunley’s presentation focused on SMET, and ignored social sciences and humanities. Dr. Bajcsy said that she and Dr. Sunley had many discussions about it, and also about the respective parts CISE and EHR could play. She spoke about a “cultural divide” between the computer scientist and the educator. Dr. Sunley said maybe there could be some joint solicitations between CISE and EHR. The AC suggested joint workshops sponsored by CISE and EHR.
DIVERSITY WORKSHOPS AND GRADUATE FELLOWSHIPS
The multidisciplinary (CISE, SBE, EHR, ENG) Information Technology Workforce (ITW) supports research focused on the under-representation of women and minorities in the IT workforce Dr. Caroline Wardle, NSF, said that women and minorities still have a long way to go in information technology. She noted that under-representation in the IT workforce is a serious national problem. The AC said that the problem is even greater with minority representation than with women and that they should be treated separately.
CISE has provided $3 -$5 M per year since 1992 on projects for women and minorities but most of the money that CISE has awarded has been put in minority institution infrastructure. Two virtual workshops were held in 1999 and summaries of both are available on the CISE web site. Dr. Wardle remarked that women’s projects are scattered with little or no theory or models resulting and that the number of women in CISE applying for and receiving Graduate Research Fellowships awards is dropping. A combined solicitation for women and minorities was issued, but until 1998 there were two separate programs. Since 1993 CISE and ENG have been putting more money into Graduate Research Fellowships (GRF) to try to increase the number of awards to women and minorities, and the number of Ph.D.s has increased slightly for women but the numbers for men are increasing faster than women.
REPORT FROM THE ASSISTANT DIRECTOR
Dr. Ruzena Bajcsy gave a status report on what’s happened since the last CISE AC meeting and said that CISE activity has focused on the ITR FY2000 solicitation, panels for preproposal and full proposal review, and preparation of the ITR FY2001 solicitation.
Dr. Bajcsy spoke about a new management approach with two types of ITR programs: base programs in each division; and ITR Initiative programs at the Directorate level. The larger grants are most likely to be multidisciplinary. A question was asked about how funds will be allocated across topic areas and how that will relate to PITAC. Dr. Bajcsy said that the first criteria will be the quality of the proposal and second, in response to PITAC recommendations. CISE’s budget request for ITR in FY2001 is for $100 M. She talked about areas in which CISE could collaborate with other Directorates such as GEO, EHR, ENG, and BIO in connection with the other initiatives. She also described meetings that were held with NIH and possible collaborations there. She addressed infrastructure issues such as investment strategy, targets, and drivers; and other issues such as diversity, the Last Mile (getting everyone connected), and the Digital Divide.
ITR SOLICITATION FY2001
Dr. Michael Lesk said that funding for the ITR Initiative may be increased by $180 M (including all directorates) in the FY2001 ITR solicitation. Dr. Lesk asked the group’s input for what new programs should be added. He said infrastructure and applications are now included. Other directorates are particularly interested in applications and in purchasing equipment for labs. Dr. Lesk said that he considers infrastructure issues such as the “Last Mile” and bringing high bandwidth to more institutions to be very important.
Examples of very large databases such as the protein data bank, astronomical data, and large collections of specimens in museums were cited as data bases that could benefit from the software research encouraged by this solicitation were also discussed. Also discussed were networking services and connectivity and whether users such as universities should pay for the services. There is also a university infrastructure and integration issue. He also talked about balancing university research with industrial research. The commercial world is making materials available for researchers to do work in networking and applications research. Dr. Lesk said there is a joint project between NSF and DARPA about collaboration and coordination in this area as well. Dr. Smith said he would like to see some connection in software research and software engineering.
Dr. Lesk said that most of the proposals are typically in the areas of software and applications. Dr. Irwin would like to see more proposals in system architecture. She said those communities are estranged, and their input should be included. Other topics suggested for the solicitation were free space technology, optical TDM, and embedded systems.
CISE INVESTMENT PORTFOLIO
Dr. W. Richard Adrion made a presentation about CISE’s role in infrastructure, instrumentation, centers, and partnerships. He addressed the investment strategy, what works best, what is the proper level of support for infrastructure and the right level for each infrastructure award. He discussed the modes, priorities, and the management. He defined the funding areas as:
- instrumentation, research resources, and facilities, both non-major research equipment (MRE);
- infrastructure, which includes basic facilities, services and installations.
He showed the level of funding for infrastructure vs. research since 1992, which originally was balanced. Over a 10-year period research has grown more rapidly and now accounts for about 70 percent of the funding. He itemized the investment strategy, the targets, and the drivers. The targets of the instrumentation and infrastructure (I & I) are fundamental IT research, application-driven research, and CISE-research-enabled I & I. The drivers enable research, provide tools, and broaden access. He said the targets and drivers change over time and the purpose of infrastructure changes as the needs and drivers change. Facilities and more infrastructures are needed to do fundamental CISE research. The issue for succeeding in the research now is the ability to build infrastructure. CISE proposals require infrastructure, instrumentation, and research resources for specific projects; and also separate instrumentation, testbed, and research resources to address general needs for research.
The questions discussed were how to link I & I to research, and how the funds should be divided in order to access the infrastructure. ITR FY2000 doesn’t supply infrastructure, and the CISE AC discussed how best to share the resources, funding, and equipment across divisions. Dr. Adrion addressed MRE-scale research facilities.
Dr. Bajcsy compared computer science today as being where physics was 50 years ago. Dr. Adrion listed some MREs for FY2001 and said that most of them have IT components and infrastructure. He asked the CISE AC to consider what CISE’s role should be. He also noted some projects that fall under CISE-research-enabled I & I that are data bases and have IT components. He spoke about the future of CISE and the existence of centers and partnerships.
NETWORKING R & D AT NSF
Dr. George Strawn gave a brief history of networking development starting with DARPA supported, CS-led networking research and testbeds from 1965-85; NSF-supported, CIO-led activities in networking systems development and diffusion from 1985-95; and NSF-supported, CIO-led high performance networking (HPN) development from 1995-2000. He said that HPN awards were made to 177 institutions. Two networks were developed (vBNS and Abilene), and assistance was provided to over 2000 institutions (small colleges) to join the Internet over the last ten years at a cost of only $20,000 per institution. At NSF, the CISE Directorate budget for FY1990 was $170 M and for FY2000 it is $388 M. The budget for Networking Research in FY1990 was $6 M and for FY2000 it is $22 M. The budget for Networking Development was $16 M in FY1990 and for FY2000 it is $44 M.
Future areas for NSF networking development is in revolutionary networking research and testbeds (to satisfy PITAC recommendations and ITR); HPN-based science and engineering (until now network developing activities such as vBNS and Abilene); HPN development (middleware); and HPN diffusion. Other disciplines need CISE leadership. The timing is right for HPN development (middleware). Dr. Strawn defines middleware as the operating system that makes networks easier to use and develops applications. We can build on the completed vBNS/Abilene/HP connections programs already in use. Research universities have been receptive to this and some may not need CISE money because of Internet 2.
HPN diffusion is popular with Congress and has a possible component of creating heterogeneous networks for testbed activities, however Dr. Strawn thinks that the timing for the diffusion may be premature because the networks are already serving universities. To get everyone involved (for education purposes) would be too expensive. Over the last 15 years, diffusion has been kept at a low expense. Typically, CISE has been very successful and active in all of the networking R & D activities, and will probably want to remain active in the near future, but the portfolio has to be balanced.
The meeting adjourned at 5:45 PM.
Dr. Baskett called the meeting to order at 8:30 AM and introduced Dr. Gwen Jacobs, University of Montana, who is the BIO AC chair.
ITR IN THE FEDERAL SETTING
Dr. Bajcsy introduced Henry Kelley, Office of Science, Technology, and Policy (OSTP) as a great proponent of computer science at OSTP. Dr. Kelley provided an overview of the administration’s R & D budget which showed that NSF has a 19 percent increase in basic research funding and an 18 percent increase in applied research funding for FY2001. For ITR, the increase is 43 percent and for nanotechnology, the increase is 124 percent. There is a $675 M increase at NSF for major initiatives and $1 B increase for biomedical research at NIH. Mr. Kelley asked the CISE AC for advice on how NIH can spend money on medical research with help from NSF CISE . There are brilliant biochemists and biologists at NIH, but they need computer scientists to advance their case.
Dr. Kelley said a team approach and the development of Centers might be beneficial. Partnerships for IT are an important feature of the FY2001 budget. Dr. Bajcsy said that although there was a recommendation for high performance computing (HPC), NIH put most of their money in the areas of human machine interface and information management. Only 5 percent of the PI’s at NIH use supercomputers but most are users of large data bases. Dr. Jacobs said she served on the BISTI Committee and the NIH community does want qualified IT people working with PI’s developing programs to solve difficult problems in biology. The idea of training centers is to get people who are cross-trained in computer science and biology.
Mr. Kelley noted the nanotechnology research themes and the increases in the budgets for FY2001 across all of the agencies. Dr. Kelley also talked about the Digital Divide and said that President Clinton recently spoke about it and challenged the community to solve the problem. The primary goal is to get every school connected, train teachers, purchase equipment, and get all students up to speed. There are several projects whose objective is to narrow the Digital Divide. Mr. Kelley also talked about educational technology and the next generation of learning technology. He is concerned that there is so little research in this area. There is a need to reform the pedagogy and teach by computer. A lot can be taught through simulation but what still must be determined is what part of education can be automated successfully. How does one get machines and people integrated for learning? The next generation of learning technology offers potential for greater efficiency.
Dr. Kelley said that the key theme of his presentation was to encourage partnerships between the information community and other disciplines.
CISE GPRA ACTIVITIES
Dr. Baskett remarked that the Government Performance and Results Act (GPRA) efforts are going well and that the measurement process is reasonable. He asked the group to review the goals.
Dr. Steve Mahaney directed the group to Tab I of the binder, “GPRA Process”, and noted that Dr. Bordogna has requested input in the form of a written report from Dr. Baskett about the usefulness of the COVs and the directorate self-assessment reports.Dr. Baskett will ask the group for help in preparing the report.
Dr. Mahaney said that there would be another GPRA report due in the fall and he recapped the process:the COV reports and the directorate self-assessment go to the advisory committees who then submit a 6-8 page report.He suggested that a draft be available at the CISE AC fall meeting. Dr. Mahaney asked the group also to review the NSF-wide GPRA report that had been distributed.Dr. Lazowska said that in reading the report he noticed that there were not many examples ofsuccesses attributed to CISE, and that CISE’s goal should be to provide examples of their accomplishments.Dr. Mahaney responded that although CISE did highlight their accomplishments for inclusion in the report, they were not adequately documented. The Director’s office has asked that they be linked to specific awards. In the future, every effort should be made to provide succinct reports.
MEETING WITH THE NSF DIRECTOR
Dr. Baskett thanked Dr. Rita Colwell, Director, NSF, for her continued support and presented the issues that the CISE AC had discussed. Dr. Colwell welcomed Dr. Baskett as the new Chair. She said that her meetings with Congress had been fruitful and she is working hard for budget resolution. She said that the CISE AC is probably the most important committee in the government in relation to the ITR initiative. She described nanotechnology as an initiative in the making and asked the CISE AC to keep pushing because it is only now evolving. In regard to women and minorities, she noted that 25 percent of the money at NSF is for education, but there is no coordination or cohesion. We must coordinate with all AC’s and communities to promote women and minorities.
She noted NSF’s good relationship with OSTP and said that CISE should take advantage of that because the workforce issue and labor shortage is a problem that can be solved collaboratively. There is a good opportunity for collaboration with NIH and with the Department of Education. Education should be seen as ongoing and seamless with an emphasis on educating students from K-12 through graduate school, focusing on advanced technologies. Maintaining data bases at a minimal cost is a challenge to the CISE AC. On the issue of collaboration with other directorates, she recognized a “cultural divide” 25 years ago, and they must work together to resolve this. The management of the ITR initiative is going well. What CISE does in this area is very important such as scalability, computation, and new paradigms. She encourages cross-directorate initiatives and asked the CISE AC to increase the volume. She asked for help from the CISE AC in preparing the FY2002 budget.
The CISE AC noted that not enough women are applying for fellowships in computer science and that there may be a bias in the review process and evaluation. Dr. Colwell said that there are many competent women on faculties and there should be some mechanism to give them money for research. She would like to have a program in place (particularly for women and minorities) which would provide their full salary, equipment, resources, and supplies for up to three years.
A CISE AC member asked how they could help Dr. Colwell to prepare for the upcoming Senate hearing. She said they should just work on promoting the 17 percent increase overall for NSF, since there is a backlog of proposals that couldn’t be funded. Dr. Baskett asked if there is mechanism for program directors to make larger awards, especially to women, for a particular proposal. Dr. Colwell said the program directors can give 5 percent of their awards for high risk projects, which they could define in any way.
INTERAGENCY DIMENSION OF THE IT INITIATIVE
Dr. Sally Howe, Associate Director, National Coordination Office (NCO), briefed the CISE AC on the responsibilities of the interagency IT RD Interagency Working Group (IWG) that is chaired by Dr. Bajcsy. She spoke about their close relationship with OSTP and OMB. She reviewed the FY2000 and FY2001 budgets by agency, and by areas for development which are: High End Computing (HEC); Human Computer Interface and Information Management; Large Scale Networking; Software Design and Productivity; High Confidence Systems; Social, Economic, and Workforce Implications of IT and IT Workforce Development; and Federal Information Services and Applications Council (FISAC).
She described the program goals as providing R&D in information technologies, using IT R&D results in Federal and national applications, and coordinating the agencies’ activities. She said that HEC accounts for 40 percent of their budget and the lead agencies for this effort are DARPA, NSF, and NSA. An important part of HEC is infrastructure and she cited several examples such as NIH’s Biomedical Information Science and Technology Initiative (BISTI). She also provided several examples of Human Computer Interface (Digital Libraries, Digital Earth) and Large Scale Networking. She said that the key players in networking are NSF, DARPA, and NIST. The common goal is wireless connectivity to the Internet for everyone, improved distance learning, and telemedicine. In the applications area, DOE is very strong in collaboration. In Software Design and Productivity, which is a new program, the NCO has been very responsive to PITAC’s recommendations. The focus of High Systems is to improve the availability, reliability, and security of software and systems, provide network security and analyze protocols, and detect and manage networks that are degraded. The Department of Commerce supports this effort and one of its focus areas is electronic commerce. NASA and DOE are the key agencies involved in Social, Economic, and Workforce Implications and Workforce Development. The goal is to develop a more skilled American workforce and to narrow the Digital Divide. FISAC provides outreach to the rest of the Federal Government agencies such as the Internal Revenue Service and the Secret Services Administration.
CISE INVESTMENT PORTFOLIO DISCUSSION
Dr. Strawn continued the discussion which began the previous day. Dr. Bajcsy asked the group whether they should take some funds that were earmarked for infrastructure and couple it with the ITR proposals or have separate infrastructure proposals. Dr. Lazowska suggested looking to universities and at the gigaflop computer community for the use of their infrastructure. Dr. Irwin talked about workshop discussions that noted that networking researchers don’t always have networks to research because they need to go outside their labs to find testbeds.One attendee said that DARPA has led that effort but there is no consensus between networking researchers and he has asked for proposals for testbeds. In response to a question that Dr. Strawn asked about whether to recommend that DARPA share NSF’s efforts in testbeds since they have been a major player, the CISE AC responded affirmatively. Dr. Baskett said that he would recommend focusing on the networking researchers, and not on applications, infrastructure, or middleware. Dr. Lazowska said that CISE needs to address supporting the best networking research possible and partnering some of it with infrastructure; facilitating access to the most advanced contemporary networks; and preparing a planned response to Congressional pressure to get improved infrastructure to universities.
Dr. Baskett said that there are a lot of companies providing middleware, so NSF would not need to get into middleware on a large scale. All agreed that NSF should not compete with industry for middleware, but in the portfolio should earmark 50 percent of funds for networking research, 40 percent for application-oriented research, and 10 percent for diffusion.
WORKING LUNCH: CISE AC ON "WHAT'S NEW AND EXCITING?"
The CISE AC members continued sharing what’s new and existing in their research.
- David DeWitt, University of Wisconsin, is working on XML data bases for large data sets.
- Mary Jane Irwin, Penn State University, is working on new ways to design computer architecture and CAD tools.
- Richard Taylor, University of California-Irvine, said that one of his students was a recipient of a software systems award and is a co-author of the Apache Web Server. The student applied what he knew about software architecture to HTTP and protocols. He also spoke about architecture description languages (ADL)
- Forest Baskett, New Enterprise Associates, talked about the progress being made in optical networks, building large optical cross connects for fibers and lambdas. He’s working on packet switches in optical networks.
BIOCOMPLEXITY FOR THE ENVIRONMENT
Dr. Margaret Leinen, Assistant Director, GEO, talked about the Biocomplexity Initiative. She said that many aspects of the environment show that lots of linear events occur with unexpected consequences and this initiative is trying to bring together many disciplines to look at them. In response to the initiative NSF established a Foundation-wide focus group with representatives from all directorates, and are in the process of establishing an advisory committee to work with the environmental community. The initiative is looking at microscale systems, ecosystems and planetary systems. The research platforms are National Ecological Observatory Network (NEON), Integrated Science for Ecosystem Challenges (ISEC), and a platform used for ocean bicomplexity research. Dr. Leinen said she is aware that research happening in CISE will change the way this environmental research has been done in the past. They are reviewing the best practices for multidisciplinary review, planning multiyear activities (5 year timeframe), identifying areas for interdisciplinary projects, and collaborating with other Federal groups. The Sub-committee on Global Change Research includes representatives from all the Federal agencies involved in the environmental initiative.
The FY2000 budget for Biocomplexity was $50 M. Over 400 proposals were received on diverse topics, and they are very anxious to engage CISE to work with them. Dr. Leinen said that climate modeling issue with its large data sets is very important.Work is currently going on to put networks under the sea to map the ocean floor and there is a huge fiber optic network there. There are many great opportunities for CISE working with GEO on this initiative.
Action Items for follow-up include:
- Provide input to CISE for Nanotechnology Program Announcement due in August.
- Advise on how CISE can work with NIH on medical research.
- Provide a written report to CISE about the GPRA process. The report should address the AC’s own experience with the process and suggest improvements.
Dr. Baskett announced that the next CISE AC meeting would be in California on October 26-27. He adjourned the meeting at 2:30 PM