Air University Review, July-August 1967
Lieutenant Colonel J. R. Brown
What changes can we expect in organizational structure as a result of advancements in automatic data processing? Will the changes evolve slowly, or can we expect abrupt shift and compliance as a result of the rapid progress of ADP technology? Will there be any dilution of middle management functions or responsibilities as a result of these advancements? These questions are prompted by recent achievement in ADP technology and its effect on the development of information systems.
Sophisticated computer-communication links capable of transferring data (or summaries and analyses thereof) on a real-time or near real-time basis may well change our thinking concerning organizational structure. Two factors are basically responsible for this change. One is the total systems concept (input or data-base oriented) as opposed to the single information flow concept (output or report oriented). The other is the improvement in computer-communication links. The input-oriented systems incorporate a broad all-inclusive data base relevant to the system and allow for extraction of these data as usable information with varied output formats. The output or report-oriented systems are less flexible because the input is limited to that which appears in the output or report format. The improved computer-communication links facilitate the processing and transfer of data on a real-time* or near real-time basis. This allows for the movement of information from source or input to the successive management levels, thereby facilitating timely management action.
There are several approaches to the subject of automation and its effect on organization. One of the more elementary approaches concerns the assignment of programmers and systems analysts. Should they be assigned to the functional agencies generating the requirements for information, or should they be under the control of the agency responsible for data processing? Another approach concerns management of the data-processing functions. Should management be the responsibility of a separate agency reporting directly to the commander, or should the data-processing functions he decentralized to several user agencies? In the event of decentralization, data processing, including computers supporting single functions, would be placed under the control of several functional agencies. Other considerations bearing on the management of data processing include the degree of responsiveness required as well as other customer needs, computer capacity, cost of hardware and software, size and location of computers. Also to be considered are two different parochial interests: on one hand, those supporting computers serving a single function; and on the other, those favoring large-scale central processors that support integrated information systems and feature time-sharing, multiprocessing, etc. Still another approach concerns the possible change in organizational structure resulting from advancements in the design and development of information systems and the speed with which information is becoming available to any given level of management. This is primarily the area on which I wish to dwell, but in order to establish a common point of departure, some discussion is necessary concerning present alignment of ADP systems.
automatic data processing
For management and control purposes, ADP systems are categorized as operations supporting, management supporting, or research and development supporting. Operations-supporting systems include command and control, intelligence, weather, etc. Management-supporting systems include personnel, maintenance and supply, financial, etc.
Systems integration. Presently we have both horizontal and vertical alignment of data-processing functions within the Air Force. An example of horizontal alignment is the major air command computer standardization program, whereby like computers are located at each major air command in support of the management data systems; another is the automated base supply system in which like computers serve the inventory management requirements at base level. Vertical alignment is typified by the intelligence data-handling system and the command and control systems, both categorized as operations-supporting data systems. These systems use computers that serve the intelligence and command and control functions at selected levels of command. Horizontal and vertical alignment applies to both the dedicated ADP systems and the mutually supporting or shared ADP systems. With the advent of the third-generation computers** and as we progress in our use of time-sharing, multiprocessing, and integrated data systems, we can foresee a possible merging of the horizontally and vertically aligned systems at the various management levels. The extent of this merger will depend largely on the considerations previously mentioned, on constraints due to the security classification of data being processed, and on the amount of systems integration obtainable.
Before any integration of command and control systems, intelligence data-handling systems, and management-supporting data systems takes place, its feasibility must be demonstrated through detailed systems analysis and design. In this instance, we must establish the degree of systems integration obtainable and demonstrate its usefulness. The hardware technology and software capability are available; the problem is to determine the degree of integration obtainable without any systems degradation.
Data systems integration between the operations-supporting systems and the management-supporting systems appears to have some practical aspects. For example, the personnel systemís combat crew subsystem and the maintenance systemís aerospace vehicle and equipment status subsystems, both part of the management-supporting systems, and the command and control systems within the operations-supporting systems utilize certain source data common to both the major systems.
Systems integration within the two supporting data systems is in-being to a limited extent. Within the management-supporting systems the procurement, supply, and financial accounting systems are integrated at base level on the supply computer. Also, studies have been made by the Hq SAC Data Systems Requirements Panel to determine the practicality of greater integration of the intelligence data-handling system and the SAC operations system, both identified within the framework of operations-supporting systems. In this instance, certain data are common to both the intelligence function and the operations plans (SIOP/EWO) function.
Centralization. Systems integration may or may not prove to be practical; however, this should not deter efforts to study the feasibility of a single, large-scale central processor with multiprocessing and time-staring features versus two or three central processors, depending upon the interrelationship of such systems as command and control and intelligence. One must also consider that several smaller-scale central processors might be as economical as a single large-scale processor and, in addition, might offer a degree of flexibility and backup not readily available with a single processor.
Centralization of data processing should result in a separate staff agency with responsibility for systems design, programming, and computer operations. This staff agency would not be a prime user of automated products and should operate as a director of information systems. However, if we retain the current alignment of management-supporting data systems, operations-supporting systems, and R&D-supporting systems, the present role for data-processing functions appears proper. One method of insuring a greater degree of control over the decentralized operation is through the use of a data systems requirements panel such as the one at Hq SAC. The panel is composed of senior officers representing the operators of the data-processing equipment and the major users of automated products. The panel does not infringe upon command or staff management prerogatives but complements normal staff action by exercising collective judgment and expertise on command-wide data-processing problems associated with new systems development, major system modifications, and hardware requirements.
effects on organization
Integrated data systems and large-scale central processors are changing the makeup and complexity of information systems. The real issue is the effect of the changing information systems upon organization. It is not so much who controls the systems analysts and programmers or who operates the data-processing center but what is happening or will happen to the structure of organization as a result of having information readily available at all levels of management. This article addresses itself more specifically to information systems incorporating, wherever practical, integrated data-processing and real-time features as well as data base orientation and inquiry techniques and their effect on organizational structure. It is apparent that most if not all routine functions of sorting, consolidating and summarizing can be effectively and efficiently. accomplished by the computer or its peripheral hardware. Optimum computer utilization, however, comes through the use of higher-level programming languages in performing the more sophisticated mathematical and analytical functions. This information is usually the result of advanced ADP systems design based on the desires and needs of management. We are already witnessing the effects of this advanced state of computer output. We are aware of the talent required to design the more sophisticated systems wherein the mass of detail data is processed into meaningful information. This, in turn, requires the exercise of exceptional talents in the portrayal and interpretation of meaningful management information.
We are observing a change in the mix of skills required to function effectively in this new and challenging ADP environment. We are witnessing more effective audit techniques and systems of checks and balances, resulting in more efficient and timely administrative action and executive control. It appears that more decisions can and will be made at higher management levels. It is at these levels that longer-range plans are formulated and that essential information is or will soon become readily available. In essence, the decision level appears to be moving up the chain of command.
Of primary concern is the development of senior executives at top management levels. The shape of the so-called top managerís ďlearning curve ď is to a considerable extent affected by his vital middle management experience. Middle management, for the most part, is staying in step with advances in computer technology, and in so doing it is able to render valuable assistance to senior executives by defining their needs and by designing and implementing meaningful information systems.
The good or bad effect of real-time systems on middle management will depend on the resourcefulness and responsiveness of middle management itself. Real-time systems will not eliminate this level of management but may dilute its prerogatives if it fails to take timely management actions. With detail data available to all management levels, subordinate levels must be especially alert to their responsibilities lest they forfeit control to higher management. The mix of skills at the middle management level will change. This change will result in fewer lower-grade personnel, offset by an increase in higher-grade personnel. The higher skills are necessary for exploiting computer capabilities and developing more sophisticated information systems as well as for programming in the higher-level languages required to support these systems.
The change in middle management may well be one of structure and composition, not dilution. Functional agencies that are involved in ADP systems development or the processing of data or that are the major users of the output of information systems will experience an accelerated change in the mix of skills required in support of these functions. With greater centralization of data-processing functions, we can expect a shift in responsibility for these functions. The greater impact will come, however, when and there is a material change in the traditional line and staff organization. Such change may not be dynamic but instead may quite possibly be reflected by an evolutionary change in the middle management structure as a result of ever improving computer-communication links and information systems.
Hq Strategic Air Command
*Paralleling data processing with a physical process in such a fashion that the results of the data processing are immediately useful to the physical operation. Robert U. Head. Real-Time Business Systems (New York: Holt, Rinehart & Winston, Inc., 1966), p.3.
**The more sophisticated computers possessing modularity to permit capacity increases or decreases. immediate-access storage, and remote input/output devices for on-line inquiry and file update.
Lieutenant Colonel J. Richard Brown (M.B.A.,
The conclusions and opinions expressed in this
document are those of the author cultivated in the freedom of expression,
academic environment of
Home Page | Feedback? Email the Editor