A. Rebrief on a specific project that is operational.

An earlier subset of what I suggest for DISA, DoD, and the Federal government is now operating in the US Army Europe. The system is known as the Information Systems Planning and Programming System (ISPPS). I provided a package on this operational system to the Suggestion panel (Dr. Signori) in the summer of 94. The point of contact for ISPPS is Mr. James Bagwell, at BagwellJ@heildelberg-emh2.army.mil.

B. What is takes to get data into the system.

Data can be loaded into DEMIS from existing list (dot-decimal paragraph numbering preferred), outline, or table files from wordprocessors, databases, and spreadsheets, which functional managers within DISA currently use to perform their regular duties. These documents can be as detailed as is available, with more detail being preferred. The intent is for these managers to use the DEMIS as a replacement for their existing methods of capturing and maintaining this information. In those cases where the managers choose not to use DEMIS to maintain this information, a periodic update would be required. The DEMIS would be out of synchronization with that data source until that periodic update occurs.

I estimate that simple lists or outlines (names and some detail attributes) of DISA locations (D4), organizations (Comptroller), work units (JTD positions) (D1), and missions/functions (D8/Comptroller) could be placed into the DEMIS with about 20 manhours of labor each.

See enclosure 1. This input would form the initial catalog of these objects (.01. = locations, .02. = organization, .03. = work unit, .04. = mission). Added to this data would be the information from the DISA Telephone Directory Service (TDS) and the crosswalk between old office codes and the new "D" codes. This would create a new type of "DISA Corporate Directory", largely satisfying a requirement from Mr. Mike Randall, now of DO3/CIO, and a portion of the DISA-IS/EIS requirement.

With the DEMIS catalog thus populated through Stage 1.1.5, it could be used as the base for moving into Stage 1.1.6. Stage 1.1.6, 1.1.7, and 1.1.8 could be accomplished through a sequential series of DEMIS survey screens (Paradox for Windows) filled out by DISA members when they log into DISANet. I estimate approximately 5 minutes cumulative per DISA member to complete these surveys at logon.

(Note that if a user cannot find the specific object they are looking for in the Catalog, they can temporarily and locally add it, with the new addition "bubbling up" throught DEMIS, to the appropriate "object proponent" to approve its permanent addition.)

This would complete Stage 1.1 of DEMIS, delivering the capabilities identified in enclosure 2.

A decision to proceed to Stage 1.2 would require that the Center for Functional Process Expertise (formerly of CIM) provide an outline or dot-decimal numbered list of known DISA, and possibly supplier and/or customer, processes with as much detail as they have avialable (Stage 1.2.1). This list could be loaded into DEMIS in about 20 manhours, thereby populating the Activity (.05.) catalog. With the .05. catalog available, Stage 1.2 could be completed by going out to DISA through another series of logon surveys, with this survey taking about 10 minutes cumulative per user.

This would complete Stage 1.2, delivering the capabilities shown in enclosure 2.

A decision to proceed to Stage 1.3. would require that Comptroller, in conjunction with the functional proponents for people, funds, information, personnel, materiel, facilities, and services (contracts), in conjunction with the DEMIS proponent, complete Stage 1.3.1. I estimate this will take approximately 30 hours per resource category. From this point, Stage 1.3 can be completed by users during logon surveys. The cumulative time for these surveys would be greater, approximately 2 hours per office, because the office is essentially building an Input, Control, Output, and Mechanism (ICOM) table for their processes defined in Stage 1.2, by picking the appropriate ICOM from the DEMIS catalogs.

This would complete Stage 1.3, providing the capabilities at enclosure 2 for that stage.

A decision to proceed to Stage 1.4 would be involve initial work by the DEMIS proponent, with little or no input from the DISA user population. This Stage could be accomplished in approximately 40 manhours, providing the capabilities identified at enclosure 2 for that stage.

Stage 1.5 would require that DISA users communicate directly with their non-DISA users, asking them to complete the same surveys which they themselves had completed earlier, up through Stage 1.3. This, in effect, brings the Customer's corporate view into DISA's management environment. The same can be done for DISA's suppliers.

Stage 1.6. moves the DEMIS from being a DISA MIS to being a DISA outcome-based, customer focused, Fee-for-Service resource management and accounting system. The managers of functions and offices could use DEMIS to capture the life-cycle quantities for resources utilized (Input, Control, Mechanism), or produced (Outputs) by their processes. The life cycle stages are: Concept, Request, Approval, Authorization, Allocations, Assignment, and Disposal. The quantities, with associated data, entered for each of these stages documents the principal information used for resource, program an project management, forming an integrated financial environment for DISA.

As the DEMIS data is refined over time, various advanced technologies can be introduced, such as artificial intelligence and expert systems (both rule-based and data-based), increasing its utility.

Stage 2 of DEMIS essentially involves using the information contained within DEMIS for a variety of structural, process, and product changes. Note that the changes made would be supported by detailed data, enabling more accurate analysis, and better informed decisions. See enclosure 3.

Stage 3 of DEMIS would find the organization operating in a dynamic, continuous-improvement, customer-focused, evolving work environment. With a modicum of discipline and maintenance, this environment can be sustained for extended periods.

C. How does the database integrate with the DISA-IS (EIS) system?

The DISA-IS (EIS) and DEMIS are highly complimentary. See enclosure 4. The EIS is designed essentially as a "front-end" to a variety of largely non-integrated databases. This front-end's role is to bring some of that fragmented information together, into a series of predefined reports, graphs, charts, etc. The DEMIS is a highly integrated database, a core system, which can serve as a "back-end" for the DISA EIS, or almost any other DISA information system. In addition, the nature of DEMIS allows the DISA-IS user, working with the DEMIS core system, to produce and save ad-hoc queries of DEMIS data, allowing them to create customized views of the organization, at any occasion, within minutes.

D. Define/Identify areas you beleive your model is suited for or to.

A point that seems to need clarification here is that DEMIS is not designed or intended as a functional application. DEMIS is intended as the basis for ALL functional applications. And the more DEMIS is used and relied upon as the system of record for DISA, the more accurate, timely, useful, and reliable it becomes.

I personally recommend that DEMIS Stage 1.1 be completed. That will provide a DISA Corporate Directory, which can replace the existing online Telephone Directory, the Comptroller's existing wordprocessing-based organization structure documents and mission and function documents, and the existing JTD Paradox application. The resultant corporate directory can be used to identify current (and proposed) relationships between functions, missions, positions, offices, organizations, and locations. This capability is currently non-existant in DISA, with each such data call requiring intensive, expensive, and highly flawed manual efforts.

Completing Stage 1.2 will provide automated support for Quality Improvement (DQIS) and other teams, a process coordination and collaboration support tool (e.g., staffing a document), a performance measurement tool, and Access Control to DISA processes, functions, offices, and locations, based on the "profile" of each position and person.

The cost for building an initial version of the DEMIS software, including deploying it onto DISANet, is estimated at $42,000. This cost is low because I have already accomplished about 90% of the actual development (analysis and design) of DEMIS over the past 12 years. I have not proceeded further in developing DEMIS because of my job's workload and the necessity of my learning a new programming language (Paradox ObjectPal) to do so. This cost estimate was provided by an IDIQ contractor (Advance) to the DISANet planners in June 94, as a cost estimate for implementing the DISA Quality Information System (DQIS). Since the DQIS is a functional extension of DEMIS, building DQIS would yield much of DEMIS. Once the software is built and deployed the data population process identified in paragraph 2 above can begin.