some comments on their use in an R&D organization
Computer Sciences Corporation
It has been more than twenty years since the first project was initiated using network techniques. Following its initial application on the Polaris Missile Development Project, networking has evolved in a direction that has enhanced its potential in assisting project management. Yet, even with the development of computerized project management network systems, the potentials of networking to assist project management often remain untapped.
The use of networking as an aid to management has met with a good deal of difficulty for a variety of reasons. However, there are two primary reasons why the potentials of networking have not fully materialized. First, there is a tendency to concentrate on adding enhancement to the technique. There are often concerted efforts for expanding the technique to encompass probability estimates, crash-time estimates, cost-time trade off analysis, branching activities, and application of optimization techniques for resource and manpower leveling. While these enhance the technique, they also result in a reliance on specialists. Removed from the process of network construction are those project engineers, task coordinators, project managers, and other personnel who should actually be using the networks as management tools. Additionally, these enhancements were not particularly useful because they generally occurred before the foundation, the fundamental procedures, and the planning disciplines on which to use networking had been fully established.
A second problem encountered in using networks is the emphasis users and managers place on the control phase of management. Consequently, very little attention is given to improving planning skills, or even undertaking planning. Networking is usually conceived of as a control technique; however, it is both a planning and control technique. It is in the planning stage where the basis for controls is established. The critical value of networking is it drives the planning process, providing the framework for systematic disciplined planning    .
Gray, Woodworth, and Shanahan  suggest another problem associated with the use of networks. They maintain that the use of networks is often discouraged by the front end costs of analyzing, constructing, and drafting of networks, in addition to the costs of collecting error-free data. One solution proposed is the use of standardized networks. Conceptually, this entails developing project models (master networks) that can be customized for individual projects .
Actually, the use of standardized networks in developing individual project networks is not a new phenomenon. Although literature on networking has rarely discussed this approach, there have probably been attempts to utilize standardized networks dating back to the initial use of PERT and CPM. The use of standardized networks discussed here refers to their use in one research and development organization (R&D). The organization is composed of 14 engineering agencies that are engaged in developing communications or weapons systems. The organization has been utilizing a computerized project management network system for a number of years. Over the last four years, the R&D organization has been using four general standardized networks — Project Master Plan models. The organization has identified its projects as having four distinct phases. The standardized networks have been developed in relationship to the four project phases. The network models developed include: 1) Concept Exploration; 2) Demonstration and Validation; 3) Full Scale Engineering; and 4) Production and Deployment.
Each model consists of activity descriptions, estimated durations, milestones for reporting, and a delineation of relationships between project activities. The models are stored on an interactive computer system. The process is to first determine which model is to be used. The model is then reviewed and revised according to the requirements of the individual project. The outcome of this process is an individualized project network adapted from a standardized network model. Importantly, the driving forces behind the use of standardized networks are two-fold. First, top-level management wanted a reporting system that would be uniform across all projects; second, developing a network model from “scratch” was considered too time consuming.
To the credit of top-level management, it was recognized that greater controls and uniform reporting were needed to increase project management productivity. Networking was selected as a technique to improve managerial productivity. Hardware was procured to support the networking function. The R&D organization's approach to networking has been to create standardized network models. The underlying assumption is that with minor revisions the model can reflect the actual composition of the project. Consequently, the focus has been only on whether or not project managers have a Gantt Chart or network diagram to model their project. How networking is being utilized to assist project management does not appear to be a widely discussed or an important issue. The use of standardized networks within the R&D organization serves to reinforce a bureaucratic and technocratic approach to management. The only consideration is that the technique is in use; the processes and procedures needed to unlock the potentials of the technique are disregarded and not fully understood.
Networking and Planning Discipline
Networking is more than merely drafting a network diagram. It is a planning discipline that includes the construction of work breakdown structures (WBS), management paradigms, systems for reporting and controlling, and network diagrams and Gantt Charts. There are four planning principles that are critical to networking. These principles include: 1) participatory planning; 2) integrative planning; 3) coordinated planning; and 4) continuous planning . The use of standardized networks negates the first three planning principles.
1. Participatory planning — The primary benefits from networking are the processes involved in producing the plans and schedules, as much as the plans and schedules themselves. Participatory planning is not to be regarded as a means of overcoming individual deficiencies. It is the means to insure that the tasks and responsibilities of project personnel occupy center stage, not the implementation of the technique. Participatory planning is critical since it provides more indepth knowledge about the content and procedures of the project, which neither standardized networks nor network specialists possess. When personnel are provided information for a greater understanding of objectives, functions, and procedures, planning sessions become more than monitoring work schedules. The sessions inevitably lead to questions of priorities in certain phases of work, potential problems affecting schedules, definition of assignments, and other details of planning that may not have been considered until these problems arose. This approach to planning possesses economic benefits since corrective actions are expensive, if not impossible, to undertake. Although this is a more time consuming process than the use of standardized networks, it does lead to other major benefits such as commitment of personnel to the plan, network technique, and project organization. Participatory planning also leads to an optimal integration of tasks, human needs, and organizational objectives and motivation for more productive work.1
2. Integrative planning — In multi-level organizations, planning occurs at all levels. Therefore, the success of the project also requires the integration of planning between levels of the organization. Figure 1 is a simplified example of a work breakdown structure. In this example, mid-level management is responsible for systems implementation. At this level, it would possess a network that encompasses preparation of computer programs, performance of tests and evaluations, development of systems documentation and procedures, preparation of systems documents and files, and acquisition of equipment. Lower-level management, responsible for the programming effort, would have a network that encompasses the development of detailed program specifications, preparation of logic diagrams and decision tables, the implementation of coding tasks, preparation of test data, testing and debugging of programs, and conducting integrated system testing and evaluation.
In this scenario, the WBS highlights the different information and planning requirements based on the level of management. It also highlights the need to integrate the planning of lower-level management with that of mid-level management. The use of standardized networks provides one network to be used by all managers without regard to the information and planning requirements of the different levels of management. Consequently, the use of standardized networks tends to generalize information requirements as if they were applicable to all management levels when they are actually applicable to a single level, or they emphasize the information requirements of one level over other levels. Additionally, because the criticality of integrative planning is not directly addressed by the use of standardized networks, there are usually inconsistencies between the level of detail of the information provided by the network and the level of detail to which the project is to be controlled. The result is that the framework for integrative planning does not evolve.
3. Coordinated planning — The complex, multi-disciplined composition of the project necessitates coordinated planning between project functions. No function can be planned effectively if it is planned independently of other functions. Without coordinated planning, projects tend to become a multiplicity of fragmented operations. The process of adapting a standardized network model to an individual project generally involves only isolating project activities to determine if they belong in that network. The use of these models seldom addresses an activity's simultaneous and interactive impact on the project's behavior.
4. Continuous planning — The project environment is constantly undergoing change. Plans that are not continuously corrected, updated, revised, and extended lose their value over time. To some extent, the use of standardized networks in the R&D organization has facilitated continuous planning. Yet, after the initial network was drafted, most of the revisions centered on changes to activity start dates, completion dates, and durations. The intent was to paint a picture that the project was under control, regardless of the project's actual state. Consequently, the planned versus actual comparisons did not reflect the real state of project control. However, the information presented from computer-generated network diagrams could usually convince top-level management the project was under control.
Standardized Networks: Advantages and Disadvantages
As indicated, the use of standardized networks in the R&D organization has been problematic. In light of their use in the R&D organization, the proposed advantages of using standardized networks have been mixed . Five proposed advantages will next be identified and discussed.
1. Eliminates the need to develop and number networks, which usually require considerable training and expertise. The use of standardized networks can make the drafting of a network much easier. In the R&D organization, the drafting of a network diagram using standardized networks took anywhere from two days to five months, depending on a variety of factors. These factors included the experience and comprehension of personnel in the use of network techniques, the experience of personnel in using a computerized system, the complexity of the project, the time constraints placed on the project activities, and the experience of personnel in using the models to draft individual project networks.
2. Significantly reduces computer input requirements; reduces input errors due to reduction of input data. Use of standardized networks does reduce computer input requirements and errors due to inputs in drafting the initial project network that serves as the baseline for planned versus actual comparisons. The computer input requirements needed to revise, correct, and update the network necessarily remain unchanged. In many projects the revisions, corrections, and updates comprise the bulk of the inputs.
3. Allows the planning process to be shifted to lower levels. There is little basis for the belief that standardized networks allow the planning process to be shifted to lower management levels. In any project, organization planning occurs at all management levels. There are differences in the approaches, however. Top-level management planning is generally concerned with the strategic development of projects in accordance with organizational objectives, formulation of project objectives, and policies for resource acquisition and dissipation. Lower-level management planning is generally concerned with the implementation of task activities in accordance with already established procedures. The use of standardized networks may tend to highlight planning at the lower levels, in the sense that planning becomes more visible. However, this is much different than actually shifting planning to lower management levels.
4. Reduces the planning time needed for each project. It is not that standardized networks actually reduce the planning time needed for each project, but they do not require any more planning than is generally undertaken. It is clear the use of conventional network techniques (e.g., Critical Path Method, Precedence Diagram Method) requires more planning than do standardized networks. This should not be considered an advantage since standardized networks do not support the systematic disciplined planning that result from the process of constructing more conventional networks.
5. Communication within and between organizations is facilitated. Standardized networks provide managers with a graphic tool to facilitate communications within the organization. Yet, the capacity to facilitate communications is also one of the strengths of network techniques such as CPM and PDM. In many regards, the process of network construction using conventional networks facilitates greater communications and comprehension than can standardized networks.
The advantages that standardized networks provide are few and relatively insignificant in comparison to the construction of networks using the procedures and methods generally associated with CPM and PDM. In comparison with more conventional network techniques, standardized networks do possess some additional disadvantages. These are:
1. Standarized networks do not distinguish between the information and planning requirements of the different levels of management.
2. They generate detailed but inconsistent information.
3. They tend to be used more for reporting purposes than for project management.
4. They are used as a substitute for systematic disciplined planning.
5. They negate the principles, processes, and procedures that are necessary if networking is to be of value to project management. Among these are the effective use of the WBS and principles such as participative, integrative, and coordinated planning.
The initial stages of a project's life cycle revolves around an environment where information is imprecise, where the activity of project managers is creative and analytical rather than administrative, where the plan is informal and not completely defined, and where procedures to be followed are unstructured and the end result is not completely specified . Networking is a systematic disciplined approach to project management whereby the ideal future is specified, the plan becomes formal and well-defined, procedures are structured, and project management becomes mostly administrative. Therefore, planning by means of networking must have the flexibility to respond and adapt to teleological variables such as goals, functions, tasks, purpose, and choice to move from an unstructured to a structured and more ordered state.
The ability to respond and adapt to teleological variables is absent from standardized networks. The use of these models in the R&D organization attempted to impose formal plans and schedules on a project without addressing the uniqueness of the work that comprises the project. Additionally, the plans and schedules developed are divorced from the knowledge and working experience of project personnel. The actual interaction and intercommunications between project personnel have not been highly valued in the R&D organization's use of standardized networks.
In the R&D organization, the use of standardized networks has only reinforced the style of project management that existed before the introduction of networking (i.e., management by intuition and crisis). The use of standardized networks is problematic because it negates the systematic disciplined planning that is the basis of networking. It is noteworthy that improved planning methodology is not among the proposed advantages of standardized networks.
After four years, no project can claim the standardized network models have assisted in the successful completion of any project or phase of a project. Most managers who have attempted to use the models in their R&D projects have found the network did not reflect the actual composition of the project. The result is managers do not possess the faith or the comprehension to use networking as a tool for project management. This is not to imply that standardized networks are of no value. In a large project organization, such as the one mentioned earlier, the value of standardized networks lies in their detailed listing and description of activities. As such, standardized networks provide a reference for planning, not for constructing and drafting an individual project network.
1. Ackoff, R.L. Planning in the Systems Age, in Ackoff, R.L. (ed.), Systems and Management Annual, Princeton, NJ: Petrocelli Books, 1974.
2. Freitag, J. Design of Systems for Planning and Control: The Systems Approach, MAPS Company, 1970.
3. Gray, C.F., Woodworth, B. & Shanahan, S. Standardized Networks: An Extension for Further Reducing Input Requirements, Project Management Quarterly, 1982, XIII, 32-34.
4. Johnson, J.R. Advanced Project Control, in Reifer, D.J. (ed.), Tutorial Software Management, The Institute of Electrical and Electronic Engineers, 1979.
5. Maciariello, J.A. Making Program Management Work—Part II, in Reifer, D.J. (ed.), Tutorial: Software Management, The Institute of Electrical and Electronic Engineers, 1979.
6. Moder, J. & Phillips, C. Project Management with CPM and PERT, Florence, KY: Van Nostrand Reinhold, 1970.
7. Sapolsky, H.M. The Polaris System Development: Bureaucratic and Programmatic Success in Government, Cambridge, MA: Harvard University Press, 1972.
8. Schmidt, K.U. Handling of Networks as Central Problem of Their Successful Application, in Ogander, M. (ed.), The Practical Application of Project Planning by Network Techniques, Halsted Press, 1972.
9. Semprevivo, P.C. Teams in Information Systems Development, Yourdon Press, 1980.
10. Standford, M.B. & Stuckenbruck, L.C. Project Planning, in Stuckenbruck, L.C. (ed.), The Implementation of Project Management: The Professional's Handbook, Reading, MA: Addison-Wesley, 1981.
Mark D. Matthews is a Senior Member of the Technical Staff, Systems Division of Computer Sciences Corporation.
PROJECT MANAGEMENT INSTITUTE NOTICE OF CHANGE OF PERSONAL INFORMATION (ENTER ONLY CHANGES TO CURRENT INFORMATION ON RECORD)
1There is no consensus on the form and content of participatory planning. This varies with the firm. Some organizations use a “committee” approach involving senior-level personnel; some organizations advocate the approach of using departmental/sectional managers; other firms approach participatory planning by having planning specialists consult individually with project personnel to develop plans. Project teams are increasingly utilized as a method to implement project planning     .