Air University Review, May-June 1983
Both Major Jones and his boss shared the perception that the MPC computer had made his assignment. After all, his career advisor stated that the computer selected him from the list of eligibles. But the advisor did not tell him that the computer had been programmed to furnish a list of officers for assignment based on their Air Force specialty code, time in service, time on station, date of last service overseas, and overseas volunteer status. However, the people in this brief scenario found it easier to blame the computer for the Air Force requirement to staff unpopular bases than to hold MPC personnel responsible for it. The computer only expedited the search and selection process previously performed by specially trained personnel, but the MPC staff determined the logic of selection and made the actual decision.
Computers have been accused of almost everything during the last two decades, from fostering unemployment to invasion of privacy in both government and industry. Alexander Solzhenitsyn recently placed the rights-of-privacy issue in proper perspective:
As every man goes through life, he fills in a number of forms for the record, each containing a number of questions. There are thus hundreds of little threads radiating from each man, millions of threads in all. If these threads were suddenly to become visible, people would lose all ability to move.1
Computers are indeed capable of tying those "threads" together, but man must first instruct them to perform the task.
Many contemporary military leaders suggest that the proliferation and misuse of computer systems are root causes of the perceived decline in Air Force leadership. I label this perception that computers control human destiny as the "Those Damned Computers" syndrome.
Air Force leaders at all levels must accept a basic truth that the use of embedded and general-purpose computers is widespread throughout the Air Force. Embedded computers are normally mini- or microcomputers that function as integral parts of larger systems (e.g., an on-board minicomputer providing radar data to an AWACS aircraft). General-purpose computers are normally "stand alone" systems characterized by management information systems, online data bases, and interactive terminals (e.g., the Burroughs B3500 computer located at virtually every Air Force base throughout the world).
Many Air Force computer systems have been misdefined, misdeveloped, and misapplied during the formative years of automation in the Air Force. On the other hand, many Air Force functional systems could not survive without the use of computers; ICBMs, satellites, payroll, etc., come immediately to mind. The purpose of this discussion is to examine the use and misuse of computers in the Air Force, analyze the failures and successes of Air Force automation, and show how effective use of computers can help leaders accomplish their mission.
A major motivation for computerization of the Air Force has been the perpetual quest by Air Force leaders for information and control of processes to facilitate the accomplishment of their missions. Information is a major source of power and a key to effective decision-making and control of processes. Computerized process control includes everything from controlling laser-guided weapon systems to robotic spot welders. Computers can facilitate executive decision-making if they are properly directed and controlled. Air Force basic doctrine states that:
The very large quantity of data acquired by our command and control systems demands many forms of data processing, so that the right information will be given to the right commander at the right time.2
The quest for decision-making information came early in Air Force history, and the quest for process control evolved with advances in weapon system technology.
Air Force leaders used computers for logistic support as early as the Berlin airlift in the late 1940s, and they have evolved steadily over the past 30 or more years. Each phase of the computer evolution was driven either by a leader’s conscious or subconscious requirement for more decision-making information and control of processes. Requests for computer capabilities to provide information or process control have taken a number of forms: Data automation requirements (DARs) for information system computers, computer resources integrated support plans (CRISPs) for weapon systems computers, or specific tasking from higher headquarters for either. Each document requesting computer power had the same basic theme: a leader either needed to acquire information or control a process in a timely manner.
Its major requirement for information and process control makes the Air Force the largest single user of computers in the federal government. In FY 82, the Department of Defense alone spent an estimated $3.6 billion on computers while all the rest of the federal government combined spends only $5 billion for computers in an average year.3 Beginning in 1983, the Air Force plans to spend more than $647 million just to replace its base-level computers, not to mention its already enormous investment in embedded computers within new weapon systems.4 Computer use has been propagated not by osmosis but by mission requirements.
Every major Air Force weapon system and virtually all management processes use computers as either the controlling or integrating function of a given system. The Air Force, indeed, is systems oriented, and computers are essential elements of most Air Force systems, from the Strategic Air Command to the Military Personnel Center.
Although improvements are necessary in many current Air Force computer systems to enhance their effectiveness, it is a truism that the Air Force could not accomplish its mission without the aid of computers. A major challenge, however, is the problem of making computers more effective tools for today’s leaders.
Many of literally thousands of computers in the Air Force are less than fully effective in satisfying requirements for information and process control. Based on 14 years in the Air Force computer systems career field, I believe that approximately half of all computer hardware and software acquisitions or development efforts fail to satisfy their original objectives or requirements. This failure, in turn, implies that at least half of the requirements of Air Force leaders for information and process control are not satisfied. The General Accounting Office (GAO) understood the magnitude of this problem when it recently reported that only 2 percent of the software developed under all government contracts could be used as delivered without changes.5 And these changes increase the cost of the system and delay the satisfaction of requirements for information or control.
A classic and, perhaps, the worst example of an ineffective Air Force computer system was the Advanced Logistics System (ALS) attempted by Air Force Logistics Command (AFLC) during the late l960s and early l970s. Some observers professed that the proverbial Captain Murphy worked on this system prior to writing his infamous "laws." The basic concept of the system was to provide on-line status and control of all AFLC logistic resources and activities. ALS development exemplified almost all of the mentioned pitfalls, and many of them occurred simultaneously. The system was conceptualized to revolve around huge on-line data bases located at major logistics centers and AFLC Headquarters. The concept was basically sound, but the development effort was beset with problems from the beginning.
The definition of ALS information requirements was not completed prior to the hardware acquisition phase; consequently, the hardware and software acquired did not match the information and control requirements of AFLC leaders. Functional users were encouraged to "blue-sky" their information requirements, and they expressed very high expectations for useful data. The Advanced Logistics System was promulgated on the basis of saving 3600 manpower spaces. The software development staff was told to develop programs for a computer "X" and to refine the software after the hardware had been selected. Thus, the entire effort to develop an on-line data base was attempted on unproven software. An obvious conclusion from the multitude of developmental problems and the knock-out punch for this beleaguered heavyweight was the lack of strong project management. After more than five years of AFLC development and an expenditure of almost $1 billion, Congress canceled the ALS effort in 1975. This massive failure did much to tighten congressional controls of all computer-related activities in the Air Force and greatly undermined Air Force confidence in developing large, complex computer systems.6
Another major DOD effort to develop a computer system has experienced severe problems and very close scrutiny and high visibility. The Department of Defense Worldwide Military Command and Control System (WWMCCS) has been under development for more than five years, and the Air Force is an active participant. To date, hardware and software expenditures have exceeded $1 billion. The objective of this command and control system is to provide information to commanders and the national command authorities (NCA) in time of crisis. A GAO report of June 1980 stated:
After almost 30 years of using computers to support the command and control environment, the Department of Defense still has not clearly defined the functional informational requirements of the military commanders who must use and rely on WWMCCS ADP capabilities for their operational needs. Instead, the Department prefers to allow the WWMCCS ADP program to continue to "evolve." It seems apparent to GAO that it is time to stop "evolving" and to begin designing a system that, when placed into operation, will provide needed support capabilities to military commanders, particularly during a time of crisis.7
Although the Department of Defense refuted the GAO comments and continued to develop the Worldwide Military Command and Control System, the criticisms are valid examples of typical difficulties facing many large-scale computer systems.
While the Air Force has experienced difficulty with "those damned computers," the problems cited are not unique to the computer systems developed by the military and the government. Private industry also suffers from most of the same pitfalls in varying degrees.8 Nonetheless, commercial businesses have at least two distinct advantages in developing computer systems; the opportunity to acquire a new hardware and software in rather short periods and relative freedom to hire, fire, or promote personnel. In the federal government, acquisition of a computer system and or termination of an incompetent employee can take years.
Another potential problem with computer systems concerns management information systems and their impact on an organization. Henry Lucas claims that organizational behavior is a primary factor in the failure of management information systems to perform effectively.9 Typical organizational problems stem from relations between users of management information and developmental staffs, lack of cooperation and acceptance by users, ambivalence to change, and potential redistribution of power within work units.
For example, organizational behavior contributes to a number of problems in the area of word processing. I witnessed the implementation of word processing systems in four different Air Force units with problems stemming from organizational behavior. Two of these organizations totally rejected the concept, and the others overcame initial problems of implementation and then enthusiastically accepted it. Major organizational problems included loss of personal secretaries working for branch and division chiefs, changes in required job skills for typists, and designation of clerical supervisors within administrative work units. People who had been personal secretaries now became mere computer operators in typing pools. The success or failure of word processing depended directly on the persuasiveness of organizational leaders and the training of administrative personnel and users. Leaders made their systems work when they recognized and understood the efficiencies and dollar savings possible from word processing regardless of the inherent organizational resistance to change.
Despite numerous instances in which "those damned computers" are less than fully effective, many Air Force computer systems prevail. Computer systems are effective if they help provide information needed by users to make correct decisions or to control resources. But computers should not be allowed to make decisions without sufficient checkpoints; they should provide information for leaders to use in their decision-making processes.
Kit Grindley and John Humble cite four fundamental indicators of effective computer systems: improved performance of repetitive tasks, increased volume of repetitive tasks, release of humans for discretionary tasks, and improved control methods.10 A good example of an effective computerized control system involving Air Force staffs was the aborted lift-off of the NASA space shuttle in November 1981. The countdown for launch was halted when one of the five on-board computers indicated a pressure buildup in a power cell. NASA officials researched the problem and discovered that the computer had correctly identified a dangerous pressure buildup in an auxiliary power unit that controlled the hydraulic system. They later acknowledged that the shuttle could have experienced difficulty during its return to earth if they had not heeded the computer’s warning. But the computer system was effective in identifying a potential problem, and NASA leaders made a correct decision based on that information. "Those damned computers" possibly saved at least two lives, billions of dollars, and the credibility of the space shuttle program."11
Another very successful computerized effort, the Joint Uniform Military Pay System (JUMPS), is one of the most effective management information systems in the Air Force. In producing more than 1.5 million pay checks per year, JUMPS is so accurate and reliable that most Air Force members cannot cite a single instance when the JUMPS computer system was responsible for an error in their pay. As a rule, human error or delays caused by the bureaucratic process are responsible. Even accounting and finance officers (noncomputer people) admit that the payroll system used prior to JUMPS was marginal at best. JUMPS personifies improved performance of repetitive tasks by effective use of computer technology.12
James MacGregor Burns speaks of executive leadership in these terms:
Classical thinking about executive decision-making has viewed the process as an essentially orderly and rational one. A problem is defined and isolated; information is gathered; alternatives are set forth; an end is established; means are created to achieve that end; a choice is made.13
Modern computers literally provide reams of information and process control for Air Force leaders. Admittedly, much of that voluminous data is not meaningful information and should be discarded, but meaningless output data are not entirely the fault of computers. Since computers are programmed by humans, effective leaders can exercise direct influence on the people responsible for controlling them, and they can influence the type of information generated or the processes controlled by computers.
Steven Alter suggests that the ultimate success or failure of a computer system is "the extent to which managers can use the system to increase their effectiveness within their organizations."14 Computer systems are successful only if they provide information that helps a leader direct or control his organization, weapon system, functional process, etc. Effective leaders will not tolerate inefficient computer systems, but they cannot improve the systems without considerable and conscious effort.
All leaders should closely examine the computer systems used in their organizations and determine whether the systems satisfy their requirements for information and control. When the systems fail to support organizational requirements, they should consult with functional users and computer staffs and seek solutions. If the systems continue to waste resources, they should either be repaired or turned off. Admittedly, the process of changing current or acquiring new computer systems is very time-consuming. But "those damned computers" will run rampant and continue to waste time and resources if leaders fail to take the initiative and demand changes.
On the other hand, leaders should exercise caution in their efforts to change standard management information systems used by the Air Force. These systems have been developed over a number of years. Many of them also have data elements or structured information based on regulations or law, and statutory requirements may prevent changes.
A key ingredient and major contributor for an organizational leader in exploiting computers is his computer support staff. Two basic understandings between a leader and his computer systems support staff are essential. First, a leader must be certain that his computer staff understands that it is first and foremost a support staff and not an independent development activity. Computer systems are not self-serving, and they should not be self-perpetuating or egocentric. A second critical understanding is that the computer support team is part of the overall team and that teamwork is paramount to the organization’s mission. The computer support staff should not feel isolated from mission areas but part of a synergistic effort. A leader should recognize his computer support staff’s positive and negative contributions to the mission and act accordingly. A staff can work wonders when members feel that they are part of the team and are recognized for their expertise and value. A motivated computer staff can wade rapidly through the bureaucratic process of acquiring a new computer or fixing a software bug. But, in many organizations, computer staffs are treated as second-class citizens and usually respond as they are treated. The treatment is justified in some instances, and, in others, negative attitudes toward the staffs are the result of "those damned computers."
An effective leader should make the best possible use of the tools provided by computer systems to gather and analyze information. As available dollars and manpower shrink, computer power can help to reduce the gap between limited resources and mission accomplishment. I believe that the U.S. Air Force’s growing pains with computer systems are similar to the U.S. Army’s problem in recognizing the value of the airplane more than 50 years ago. Today’s leaders must learn to appreciate and exploit modern technological advances. With the assistance of their support staffs, they should refine their baseline computer systems and use them as building blocks for enhanced capabilities.
The refinement process should begin with a conscious appraisal of the what, when, where, why, and how of computer use throughout the entire Air Force. To ensure objectivity, a disinterested third party or a contractor should make the assessment and brief the Air Staff and all major commands on the results. These actions would mark the beginning of an Air Force-wide program to make computers more effective. Major aspects of the program would be identification of effective and ineffective computer systems, necessary actions to enhance current effective systems, redirection or cancellation of ineffective systems, and identification of functional areas or processes that offer potential "high pay-back" for the use of computers.
Another much-needed initiative is to focus attention on information as a major resource and manage it as carefully as we manage logistics, personnel, dollars, etc. Computer and communication systems are technical solutions to information requirements and should be viewed as subsystems in a total information systems approach. The recent Air Staff initiative to form an Assistant Chief of Staff (ACS) for information systems is a positive first step in developing an information resource management approach. However, promulgation of this concept throughout the Air Force is a major undertaking that will involve fundamental changes in the way the Air Force manages its computers and communications.
A study by the Rand Corporation in 1976 summarized the importance to the Air Force as follows:
Computer technology has become central to the Air Force’s ability to perform its role and mission. Computer hardware failures and inadequate software can mean that an aircraft cannot fly or cannot carry our its design missions, that command and control systems cannot communicate with each other, or that important management information is nor received in a timely or accurate manner. 15
Since that study was completed, computer use in the Air Force has continued to grow at an exponential rate. Whether "those damned computers" run the Air Force or whether they are used effectively to satisfy mission requirements depends heavily on teamwork among computer users, support staffs, and strong leadership.
Air Command and Staff College
Maxwell AFB, Alabama
Notes
1. Donald H. Sanders and Stanley J. Birkin, Computers and Management in a Changing Society (New York: McGraw-Hill, 1980), pp. 1-137.
2. Air Force Manual 1-1, United States Air Force Basic Doctrine, 14 February 1979, United States Air Force, p. 2-24.
3. Robert V. Head, "The Complex Nature of Federal Data Processing," Government Executive, March 1981, pp. 30-31.
4. "We Have Play-offs, Runoffs, Bake-offs; Now a Compute-off," Wall Street Journal, September 21, 1981, p. 1.
5. Robert V. Head, "Federal ADP Systems: Atrophy in the Sinews of Government," Government Executive, February 1981, pp. 36-41.
6. Air Force Logistics Command, Advanced Logistics System Assessment, Executive Summary (Wright-Patterson AFB, Ohio, 1975), pp. 1-53.
7. General Accounting Office, The Worldwide Military Command and Control System—Evaluation of Vendor and Department of Defense Comments (Washington: Government Printing Office, 30 June 1980), pp. 2-3.
8. Kit Grindley and John Humble, The Effective Computer (New York: McGraw-Hill, 1973), pp. 1-18.
9. Henry C. Lucas, Why Information Systems Fail (New York: Columbia University Press, 1975), pp. 1-30.
10. Grindley and Humble, pp. 21-60.
11. "Shuttle Launch Delay Attributed to Oil, Software Problems," Aviation Week & Space Technology, November 9, 1981, pp. 20-21.
12. Barbara Englert, "Air Force Pay Service—The Best from the West," Air Force Comptroller, July 1981, pp. 8-9.
13. James MacGregor Burns, Leadership (New York: Harper and Row, 1978), pp. 369-97.
14. Steven L. Alter, "How Effective Managers Use Information Systems," Harvard Business Review, November-December 1976, pp. 97-104.
15. Stephen M. Drezner, Hyman Shulman, et al., The Computer Resources Management Study (Santa Monica: Rand Corporation, 1976), pp. 1-29.
Contributor
Major H. Wayne Wolfe
(B.S., Otterbein College: M.C.S., Texas A&M University) is Chief, ADP Plans and Programs, Hq Pacific Air Forces, Hickam AFB, Hawaii. He has served as a computer programmer, systems analyst, operations officer, and executive officer at various Air Force data automation activities. Major Wolfe is a graduate of Squadron Officer School and Air Command and Staff College.Disclaimer
The conclusions and opinions expressed in this document are those of the author cultivated in the freedom of expression, academic environment of Air University. They do not reflect the official position of the U.S. Government, Department of Defense, the United States Air Force or the Air University.
Air & Space Power Home Page | Feedback? Email the Editor