It is common to find that project owners, contractors and managers react to prospective schedule overruns by adding resources to try to make up time. This behavior leads to more cost overruns as it moderates (hopefully) the schedule overruns. Projects that are schedule driven will experience this phenomenon acutely as the owners of the finished project have been promised a date. Also, any project needs to start generating revenue in order to be economically viable. Finally, for NASA, promises made to Congress include estimates of both time and money but schedule is increasingly important. Any serious delay jeopardizes the relationship with Congressional committees by either overrunning the schedule, descoping the final product or both. There is no guarantee that adding resources will have the desired effect – schedules may not be recovered, but may h less overrun than otherwise without the resources.
Typically, analysts model the current plan as if the manager will not react if the schedule is jeopardized. This is not realistic. When simulating a project schedule the analyst needs to have a way to represent the manager’s response to delays at the beginning or during execution, when there is still time to preserve the schedule.
Conditional branching can represent the project manager’s response to a schedule event such as the detailed engineering’s ending later than anticipated. Unless this can seamlessly be translated to a schedule overrun without penalty, something must be done. With conditional branching, during the Monte Carlo simulation of the project schedule, a test for missing a detailed engineering finish date is performed in each iteration. Depending on that test – finish date of the detailed engineering -a branch of the schedule that models the management decision as anticipated is chosen. (1) if detailed engineering is finished not too late, then the project continues with the current plan. (2) if detailed engineering finishes too late, a “recovery plan” is adopted with shorter durations for construction / execution and the associated increase in resources needed to make that new plan a reality. Using conditional branching the expected values for cost and schedule taking into account management actions can be simulated.
