Cybernetic Control – Project Control Techniques

Cybernetic Control – Project Control Techniques (#1 in the series How to Control a Project)
By Michael D. Taylor

Having a project management plan will not always ensure having effective project control. Without a control process the project manager will often resort to an improper use of institutional authority to embarrass, or intimidate a project member whose performance is unsatisfactory. As a result the project member will learn to prevent disclosure of any problems. This then creates another problem in that the project manager is not being made fully aware of deviations from the project plan. Taylor’s Law1 states that “the earlier a problem is disclosed, the easier it is to manage.” When project problems are hidden from the project manager they often grow to the point where they become untenable.

Meredith and Mantel offer three methods of control, these are:

• Cybernetic Control
• Go/No-Go Control
• Postcontrol

Cybernetic control involves a self-correcting feedback loop as illustrated in Figure 1. Project control can be defined as the ability to identify deviations from a plan or a baseline, and take corrective action to get back on course. Therefore, in order to maintain cybernetic control of a project, four vital elements must be present.

Figure 1: Cybernetic Control Process

First, an accurate, agreed-upon baseline must exist. The baseline consists of the project schedule, the project budget, and any other measurable aspect of the project. The initial baseline is established toward the end of the planning phase when the Final Planning Review is conducted. If all planning conflicts are resolved at the FPR, an accurate, agreed-upon baseline should be established. This effectively becomes like a compass against which a subsequent performance is measured.

Second, a means of measuring actual work is required. This can be a challenge due to subjective estimates as to how much of a given project activity has been completed. Assigning a percent completed to a given project activity is often a guess on the part of the individual who is responsible for that activity. In some cases, the individual may give an overly optimistic estimate of how much of the activity has been completed in order to avoid embarrassment during project status meetings. This phenomenon is often prevalent when the project manager uses “negative means” of controlling variances from the project baseline. To avoid this dilemma, techniques such as the Earned-Value Management weighted milestones method can be adopted.

Third, the control system must be capable of identifying variances from the baseline. Most project schedule applications will provide variance reports provided it has the baseline and actual work information. The software will determine the differences between these two factors and will provide variance information to the project manager. It is these variances that are typically reviewed during the regular project status reviews. The control chart below illustrates how variances can be identified using the Earned-Value Management (EVM) method.

Figure 2: Project Tracking Control Chart Example

Fourth, corrective actions must be taken for those variances that are deemed unacceptable. Variances in excess of 10% should be considered significant, and the project manager should require the “owners” of the unacceptable variances to develop and present the steps needed to get back to the baseline plan as quickly as possible. Either “soft” or “hard” recovering planning must be conducted. This will be described later in this series.

MICHAEL D. TAYLOR, M.S. in systems management, B.S. in electrical engineering, has more than 30 years of project, outsourcing, and engineering experience. He is principal of Systems Management Services, and has conducted project management training at the University of California, Santa Cruz Extension in their PPM Certificate program for over 13 years, and at companies such as Sun Microsystems, GTE, Siemens, TRW, Loral, Santa Clara Valley Water District, and Inprise. He also taught courses in the UCSC Extension Leadership and Management Program (LAMP), and was a guest speaker at the 2001 Santa Cruz Technology Symposium. His website is www.projectmgt.com.