CENDI PRINCIPALS AND ALTERNATES MEETING

National Agricultural Library
Beltsville, MD
December 7, 1998

Minutes

FCW IT Industry Forecast: Markets and Developments
NRC Computer Science and Telecommunications Board: Recent Studies and Future Plans
EPA Overview and Plans for STI Management
Recent Development in AgNIC and AGRICOLA

WELCOME

Tom Pedtke, CENDI Chair, began the meeting at 9:10 a.m. He thanked NAL for hosting the meeting. Introductions were made. Since this was the first CENDI meeting at NAL, Ms. Andre introduced NAL and its facilities.

TECHNOLOGY AND POLICY FUTURES

FCW IT Industry Forecast: Markets and Developments
Edith Holmes, President; Steve LeCompte, Vice President for Research, Federal Computer Week )

Federal Computer Week (FCW) is part of the Government Technology Group/International Data Group, that publishes InfoWorld, MacWorld, and PCWorld. FCW is twelve years old. During this time, it has broadened its scope from federal IT to state and local government IT, and from print only to Web dissemination (civic.com and fcw.com). Every two years, FCW holds a Government CIO Conference. This September, they also held the Advanced Information Technologies Conference. A major trade show in partnership with Armed Forces Communications & Electronics Association (AFCEA), called TechNet, will be expanded next June and called GovTechNet. CENDI will be providing a program for this conference.

Mr. LeCompte and Ms. Holmes discussed the federal information technology issues, spending, infrastructure, and staffing. The presentation was based on surveys of the FCW readership and of special groups within the government community.

The primary focus is on the Chief Information Officers (CIOs), since they are now the focal point for IT. The main issue for the CIOs is the requirement to show a return on investment similar to a private enterprise (OMB/Raines Memo 10/96). The CIOs are trying to determine how to calculate return on investment (ROI) in a government environment.

The CIOs are currently driven more by tactical and oversight requirements of the Government Performance and Results Act (GPRA) than by long-range issues. The main concerns among both the civilian and defense federal communities are IT capital planning, Year 2000 compliance, and enterprise management (network resource management). The defense community is also concerned about the cost of ownership across the life cycle of the hardware and software. The educating of politicians is of more concern to the civilian agencies than for defense.

According to OMB budget data, the Executive Branches' IT market has increased from approximately $9B in 1982 to an estimated $30B in 1999. Information Technology spending per employee is increasing quickly, while the number of employees is decreasing. However, the CENDI members indicated that the growth is not reflected in the IT budgets of the STI programs. Much of the increase is going to financial and personnel systems as well as Y2K compliance.

Year 2000 programming/consulting services are also reflected in the growth in the expenditure in the services area. As agencies move to outsource and lease equipment rather than buy it, the increase of expenditures on support services is likely to continue. The purchase of seat management services may also be causing the increase in the services area.

Unfortunately for the analysis of federal IT spending, OMB has stopped collecting a lot of the detailed data from the agencies. However, some information is gathered by FCW at its conferences. At the last CIO Summit, they asked about the system constrictions. The responses indicated that there is a great deal of end user demand for the latest client hard/software capabilities.

The purchasing methods are changing, with the traditional agency and other "indefinite delivery-indefinite quantity" procurement vehicles being replaced by GSA. There is also an increase in credit card usage. These changes are corroborated by industry.

The shift from the mainframe to LANs and WANs (the Intranet and Internet) is well underway. It is surprising how many mainframes are still around. Despite the inability to keep up with the demand for desktop clients, the rate at which PCs are being introduced appears to be flattening out.

In the federal government, 486 PCs are being upgraded quickly to Pentiums and Pentium IIs. There is a significant move to Intel-based desktop machines running under Win NT, rather than RISC under UNIX. The Macintosh is not heavily used except for some locations at NASA and DOE. The result is that the federal infrastructure is relatively modern. The refresh rate (the rate of replacement) is accelerating to every four to five years. Procurement reform has caused faster refresh rates. Some industries are reporting that the selling cycle to government is now shorter than for private industry.

Based on an October 1998 survey comparing the ratio of users who plan to buy a product to those who already own the product, the hottest products next year will be Y2K services, Java, and Virtual Private Networks.

A study of security elicited no strong responses. CIOs believe they are doing the best in the area of logical security, followed by physical security. The weakest area is the education and training of staff to be concerned and aware of security issues. Privacy is really driving the security issues at this time, rather than issues of national security. CIO opinions about privacy versus security are very ambivalent. On the one hand, they believe that privacy is critical to our form of government; on the other, they believe that openness is critical to freedom. There was strong disagreement when asked if privacy should be sacrificed for the benefit of electronic commerce. Mr. LeCompte sited the problems with the Social Security Administration (SSA) earnings data on the Web. They had not thought through the issues related to individual privacy, and the service was taken down within hours because of media attention and public perception. Electronic commerce is at the bottom of the CIO concerns, perhaps because other issues have such high profiles. Also, there is no consensus on who should run electronic commerce, and no federal agency has volunteered to take it on.

On the horizon for the federal IT community is "Seat Management", the acquisition of desktop and network computing services as a utility on a per "seat" basis. Services are being treated as utilities and products as commodities. NASA and GSA have announced big projects related to Seat Management. The goal of these projects is to achieve cost savings and maintain technological currency by making it the responsibility of the contractor. This approach automatically standardizes the IT environment. There is nothing solid yet to show that this approach actually costs less than leasing or owning. Ms. Holmes believes that it is likely to be at least the same or perhaps more costly. Seat management can be difficult because IT professionals may not understand service-level contracts. The annual budget cycle also causes problems for contractors when the budget cycle requires that risks be taken by the contractor. Seat management emphasizes the programs and mission of the agency rather than the hardware/software. This may make it a more positive move in the long term.

Summarizing the federal IT infrastructure, Mr. LeCompte indicated that MicroSoft is clearly in the driver's seat for both client and server software. Servers will be "hot spots" in the future, particularly those that are public, because they will define the visibility of the agency.

A look at the federal IT workforce shows that while downsizing has hit many other jobs within government, it does not seem to have effected the IT worker. The approximately 110,000 IT workers are about evenly split between defense and civilian agencies.

Keeping the IT staff up-to-date is a challenge. A great deal of learning is required each year, but CIOs and the staff disagree on how this should be accomplished. CIOs prefer formal training, while informal self-instruction is preferred by the technologists.

Discussion

Dr. Warnick mentioned the need to be better connected with the CIOs. In some CENDI agencies, such as NASA, the STI Programs are now reporting to the CIO. This is not the case with all. It is difficult to get STI information management concerns addressed because the CIOs are focused on other things. However, all agreed that the CIOs need to better understand the content management issues.

NRC Computer Science and Telecommunications Board: Recent Studies and Future Plans
Dr. Jane Bortnick-Griffith, Interim Director, NAS/NRC Computer Science and Telecommunications Board

The Computer Science and Telecommunications (CST) Board of the National Research Council (NRC) is a combination of academia and industry. It monitors and promotes the health of computer science and technology in order to fuel future developments in these areas. There is emphasis on core research. It initiates studies and responds to requests for studies from other government agencies. Some are funded by agencies such as NLM (medical applications on the Internet) and Defense Advanced Research Projects Agency (DARPA), (command and control systems), while others are mandated by Congress in legislation and then funded by the agency (limiting pornographyCtechnologies for the Department of Justice). The NRC itself is the operating arm for the National Academies of Science. The CST is part of the Committee on Physical Sciences, Mathematics and Applications.

The CST Board meets three times a year. Its board members are volunteers. The benefit gained by using the NRC is that it is a peer reviewed, neutral body. The agency requesting the study has no substantial input into the study while it is being conducted. Other than a complimentary briefing before the results are published, the agency does not have any involvement in determining the results.

The CST is conducting approximately six studies at any given time. Some of the current and recently completed studies include:

Finally, three more reports are due out in February:

The cost of a study varies depending on its duration and requirements. Multi-year studies range from $500,000 to $800,000.

Dr. Bortnick-Griffith provided a review copy to the Secretariat of several reports and then took names of those interested in receiving copies. Information about the reports and some draft texts are available from the NRC Web site.

Discussion

Dr. Warnick asked about the interaction of the Board, the NRC and the academies with relation to high level issues such as intellectual property. He cited the request for extension of the PubMed capabilities to other disciplines and the letter to the legislators regarding the Database Protection legislation. Dr. Bortnick-Griffith indicated that high level issues, such as intellectual property, are done by the academies themselves, rather than at the NRC or committee levels.

EPA Overview and Plans for STI Management
Dr. Robert Shepanek and Dr. Gary Foley, SIMCorB, EPA

Dr. Robert Shepanek and Dr. Gary Foley of EPA's Office of Research and Development (ORD) were introduced. Dr. Shepanek is with the National Center for Environmental Assessment. Dr. Foley is Director of the National Exposure Research Laboratory.

Recently, the EPA administrator determined to change how the Agency is organized to manage environmental information. The basic premise underlying the reorganization is that combining management of information technology (IT) with management of information content will be a more effective way to support many of EPA's strategic objectives. These objectives include a citizens "right to know" EPA information, integration of environmental data from a variety of sources, and reduction of the reporting burden on the regulated community through more effective use of IT. Ultimately, through the re-organization, EPA will do a better job of providing environmental information of known quality for secondary use, and mine more from the data that is collected. The reorganization will better position

the EPA CIO to integrate information policy and collection, information technology services, and information analysis and access.

Along with these changes in the approach to environmental information management at the Agency level, the EPA Office of Research and Development (ORD) is altering the way it approaches the management of scientific and technical information. ORD has recognized for some time that an evolution in the practice of environmental science is taking place. Environmental research and monitoring efforts are being conducted at larger and multiple scales. Environmental health and ecological assessments require use of data from disparate sources. Fewer resources are available for data collection making the "secondary" use of data more important. In order to meet the challenge posed by these changes, IT needs to be better leveraged by the environmental science community to carry out the collaborative efforts required to accomplish the science.

In order to use IT more effectively, EPA/ORD began by developing an Information Resources Management Strategic Plan for the Office. The Plan called for the charter of a high level information management coordinating group with representation from all of the ORD laboratories and centers. The Scientific Information Management Coordination Board (SIMCorB) was created. Its focus is on the development of both the organizational and architectural aspects of a comprehensive computational environment for ORD. The SIMCorB effort is making more effective and efficient the collection, analysis and integration (including the provision of interpretive results), quality control, maintenance, dissemination, sharing and archiving of ORD scientific data, information and tools.

The Board itself is a virtual organization that will not have its own permanent staff. An executive advisory committee provides high-level management review of SIMCorB plans and initiatives and is populated with representatives from organizations throughout EPA. The SIMCorB effort has been well received, but resources for IT within ORD remain a significant issue.

One example of SIMCorB initiatives is the Environmental Information Management System (EIMS). ORD, other EPA Offices, EPA Regions, and States are using EIMS for the distributed creation and access of data, metadata, etc. EIMS uses web technology (Oracle and OracleWeb Server). EIMS can be accessed at www.epa.gov/eims. The system is assisting ORD and EPA in describing and relating multiple types of information objects such as projects, documents, databases and models. Lotus Notes/Domino is being used by ORD for document management.

A Five-Year Science Implementation Integration Plan is being developed. Dr. Shepanek asked CENDI to perform a peer review on the plan when it is completed. In addition, SIMCorB is coordinating proof-of-concept projects and developing its relationship with the new EPA Information Office as well as its relationships with external organizations. Its involvement with CENDI is part of the latter, along with increased involvement with the National Biological Information Infrastructure (NBII) and the National Spatial Data Infrastructure (NSDI).

Recent Development in AgNIC and AGRICOLA
Deborah Richardson and Melanie Gardner, NAL

The NAL Web Site is divided into many separate areas. Ms. Richardson took the members on a brief tour of the site, emphasizing items of interest such as the Web Policies, the Personnel Directory, and General Reference Information that provides links to other Internet resources. There are several special image collections, including Rare Botanical Prints, Plant Pests and Diseases, and a large collection of black and white Forest Ranger photos.

Part of the Web site are the contributions from the Centers of Expertise, NAL's information reference centers which specialize in certain agriculture topics. They provide expert reference in these areas and unique physical collections. Specialized areas include Animal Welfare, Food and Nutrition, Genetics, etc. The information centers answer about 50,000 questions annually. There are also 160,000 document delivery requests fulfilled. Each center has developed Frequently Asked Questions to help in the answering of questions. Questions that are out of scope for one center are routed to the appropriate center.

NAL's 3.5M records are available to the public over the Internet. The site is divided into books that are available from NAL's library Online Public Access Catalog, and the journal literature from the AGRICOLA file. The NAL site received 10M hits during 1998, but it is difficult to tell how the availability of AGRICOLA will change this, since the statistics include only one month of AGRICOLA Internet availability.

The Agriculture Networked Information Center (AgNIC) is a Web-based service originally developed in cooperation with the Coalition for Networked Information in 1994. NAL has since taken over primary stewardship for this system. They provide the Secretariat but it is an independent organization. The alliance has grown from five to twenty partners, with six more waiting to join. Governance documents, available on the Web, outline the way the alliance works and the responsibilities of the partners. Oversight is provided by an Executive Board with NAL providing the Secretariat.

AgNIC includes a variety of services. AgDB is a catalog of over 900 evaluated sites/databases available on the Internet. NAL and its partners create a calendar of agricultural conferences and meetings. AgNIC also links to over 200 calendars maintained by other organizations. There were over 70,000 hits on AgNIC per month. Seventy percent of them are hits against the AgDB and Conferences files. NAL plans to connect the conference calendars to the electronic proceedings from the conference.

The Web site is currently being overhauled, including the subject categories used to organize the site. The Executive Board is working on further stabilizing the alliance.