Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
UNITED STATES AIR FORCE PROJECT IMPACT John Zaner Air Force Systems Command I would like to begin by giving you a little of the background of Project IMPACT. The primary motivation behind Project IMPACT was a study conducted by the Air Force in the l975-76 time frame. This study pointed out that although the Air Force work load was going to be increasing, particularly in terms of information gathering and dissemination, the workforce to support that collection and dis- semination process would decrease or at best remain constant. Faced with that situation, Air Force Systems Command established Project IMPACT in l978 to investigate the possibilities of using office automation to improve productivity in the information use area. Its objective was to modernize systems management and systems acquisi- tion operations in Air Force Systems Command, but with an under- standing that the cost of the implemented tools had to be offset with gains and productivity of the clerical and professional work- force. In initiating Project IMPACT, our past experience indicated that many of the successes or claimed successes in office automation had not really reached the potential that was forecast for them. In fact, failures were numerous and expensive. This was true, not only in industry, but in the government as well. One of the biggest problems appeared to be a lack of acceptance by the user. This was both from the standpoint of people not wanting to change the way they did things and from the standpoint of having systems that were not oriented to the user. There was a perception that office auto- mation was difficult to use and therefore not acceptable. Further, the cost of what we proposed to do was not really defined. There- fore, we found it was very difficult to measure increased producti- vity when we did not really know what we had in our current manual system. So, our general conclusion a couple of years ago was that we should move very cautiously in this area and attempt to have a lot of people involved in it, especially users and implementers. 49
We established Project IMPACT consistent with our overall pro- gram management philosophy. We set up a program office to manage the program. We hired a prime contractor, Booz-Allen & Hamilton, to support the program office and, we are conducting it as a multi- phased project. The first phase is to evaluate what is going on in government and industry in terms of office automation. The second phase is to conduct extensive interviews and select candidate systems for prototype testing and candidate hardware for implementation. The third phase is to design a prototype for implementation. Phase four, is to implement and test that prototype. This is the phase that we are in now. The fifth phase is to document the results of the prototypes and measure them against savings that we projected in Pnases one and two. One of the interesting aspects of IMPACT is that we chose to use a product orientation rather than a functional one. Within our field organizations there are a number of products that are produced over and over again. For example, if we are going to acquire a weapon system we need a request for a proposal, we need a statement of work, and we produce any number of technical documents in support of that weapon systems acquisition. So, we chose to look at each product and follow it through the flow of the organization, identi- fying those which were labor-intensive and those where there was a high pay-off in using automation. We collected information concerning the products by conducting extensive interviews from the commander right on down to the working level. Based on the interviews and some measurements, we establi- shed the amount of time that went into the creation of each of the various products in terms of the tasks supporting the product and the functions that were performed in support of producing the products. We got some very interesting results. First of all, and it didn't surprise a lot of people, the number one product in the Air Force Systems Command is briefing charts. It is the way we run our organization. The Air Force Systems Command management style is based on briefings all the way up the chain of command. Based on the results of the inteviews, and using a computer system to help us reduce and analyze the data, we prepared a product oriented list which shows the greatest consumer of man-hours down to the lowest consumer of man-hours. When we originally did the study we came up with some 220 products. We couldn't fully analyze that large number of products so we conducted another series of interviews to identify those that were really the most critical. 50
Chart 1 represents some results of our examination of one of our major organizational elements/ what we call a product division. We have six product divisions in Air Force Systems Command: Ballistic Missile Division, Space Division, Electronic Systems Division, Armament Division, and the Aeronautical Systems Division. The results of our analysis produced figures which were fairly constant across all product divisions. This particular set turned out to be most interesting. The first column (Ct) represents the percentage of total hours of clerical people doing clerical things. The second column (Pr) represents a professional person's total work activities. The third column (PCt) represents a percentage of the value (Pr) which a professional person spent doing clerical things. Some of the products on Chart l may not be familiar to every- body. The briefing is, in fact, a briefing much like the one I am giving today. In terms of briefings, this product division produced l7,020 a year. The total number of hours that went into the prepara- tion of briefings represented l20,000 man hours a year. The second line is the time to prepare and process a contract through the procurement snop. The third item is for travel orders. We found that it costs about $90 a shot to prepare a travel order and yet they serve only a minimal useful purpose. We spend a phenomenal amount of money preparing travel orders and everybody agrees that we should get rid of them. Consistent with this type of finding if there is no use for a document, we intend to get rid of it. The next two high interest items on the Chart are the preparation of specifications for a contract, and the development of integrated logistics support plans. Wnile reviewing our findings, we have found that some other interesting conclusions can be derived from the numbers. For example, if we look at the current labor utilization in this parti- cular product division, half of the time that went into products is professional time. The other half is clerical time. Even where it is a professional doing the clerical work, the mix of work between professional and clerical is 50-50. If you look at our current mix of people, we have four professional people to every.clerical person. So, there may be something wrong with the figures, or maybe our esti- mates in the past of how many people a clerical person can support are erroneous. We are implementing a prototype system to prove whether or not such hypotheses are correct and whether it is pos- sible through office automation and local management information system (MIS) kinds of techniques to recover 47 percent of the clerical time spent by clerical people and l8 percent of the time spent by professionals doing clerical work. We believe these figures to be conservative. We have enough evidence to back us up 5l
that we can make those kinds of projections and feel very comfort- able with them. Of course, we will know a lot more by December or January when we start getting some of the actual results of the prototype implementation. However, some of the recommended improvements are so obvious that we are not waiting for the prototype. For example, the pre- paration of briefing charts: it is obvious we can increase produc- tivity through some sort of a graphics processing system. So, we have planned, and are now involved in implementing, a centralized graphics processing system in our products divisions that can eventually be integrated with the office automation function. The elimination of travels orders is another obvious recommendation. What will the prototype look like in a product division? In each product division, we have the commander, his various deputies, who are like vice presidents, and then under the deputies a number of special project offices. The latter are special offices to acquire a major weapons system, such as the B-l, or the F-l6. Each is a self-contained team, set up for that particular acquisition. The program manager, who heads the team, has the management and functional people necessary to support the acquisition of the particular system. What we will provide the program manager is a distributed capability implemented within his program office for text editing, electronic filing, some very simple management tools, like managing a schedule, some local information retrieval and very limited data processing. In fact, that kind of capability that can be implemented on a fairly complex word processing system. At the deputy level, we will centralize some more sophisticated MIS capabilities and some product functions. Each system will augment the user. For example, if you want to prepare a contract, a tutorial system would lead you through the preparation of a con- tract. The procurement specialist sits down at the terminal and says, "I want to prepare a fixed price contract." And, the system goes through all the various clauses that are candidate for that contract. He can strike out information that he doesn't need. He can supplement it at any point. That part of the prototype has been implemented and it has been very well received. The particular installation looks something like the simplified portrayal on Chart 2. The implementation at the deputy level is on a large minicomputer, the one we are using in the prototype is a Prime 550. One of the most significant aspects of the prototype implemen- tation is that we are making maximum use of the vendor supplied data base management system rather than develop our own unique software. 52
Within the program offices we have several word processing systems. One of the side issues involved in the prototype implemen- tation is to evaluate word processing systems that are available on the federal supply schedule. To do this we have seven different kinds of hardware and 30 software systems installed. We have evaluated at least one of each of all the word processing systems that are currently available on the GSA schedule and are collecting information on user reaction, environmental considerations, and other related items. IMPACT is looking at office automation, anything that can improve productivity in an office, from word processing to executive telephones to electric letter openers if that will improve producti- vity. We also have another program, the Command Management Informa- tion System (CMIS), which views IMPACT as a source data automation mechanism for collecting management information, rather than have people in the office environment use a word processing system for preparation of correspondence, electronic mail, and then turn around and feed some other kind of terminal into a management information system. Some of these people currently have four or five different terminals in their office, one for each different kind of MIS that they need to feed. We want to start integrating them and tap the information out of the word processing systems at the source. That, and the integration of information up through the chain of command to the commander, are the major emphases of the CMIS program. In that program we currently have an initial operating capability where our commander. General Slay, has direct access to some decision support capabilities that are coming in from the program offices. In the future, we see a combination of the Command Management Information System and IMPACT. Each has their separate place in the world and we are currently treating them separately during the study phase of IMPACT. However, CMIS will provide the implementa- tion vehicle for IMPACT probably next year. I thank you. 53
CHART 1 Quantitative Results TT Hrs/ Prod. Total Hrs/ Prod. Percentage Pro No. Yr. CT/TT PT/TT PCT/PT Briefing 1,720 70 120,000 .34 .66 .45 Contracts 85 945 80,000 .31 .69 .06 Travel Orders 16,000 4.2 67,000 .76 .24 .33 Specifications 55 1,126 62,000 .40 .60 .06 ILSP 50 1,240 62,000 .19 .81 .18 CHART 2 Project Impact Test Configuration Large Mini Applications Storage Translation Seven Generic Types 30 Units WP 54
