Project Management Learning Sim

For NASA, Ventana developed the Project Manager Learning Sim: a training tool containing a simple project-dynamics model, giving users access to the model via a game-like user interface. The Sim was developed in order to enable users to discover for themselves several fundamental insights into managing real world projects, the most fundamental of which is that progress data is critically important to managing a project successfully.

The Sim represents a project to build the first prototype of a launch vehicle (LV). The Sim’s main window (Figure 1) shows information on actual and planned expenditures (on the right side of the main window) as well as on design-adequacy (left side).

Figure 1: The Sim’s main window shows information on design adequacy (left) and expenditures (right). The “add data feed” button in the lower right provides access to additional information.

The fundamental design “trick” is that additional data feeds are available, but are not provided on the Sim’s main window. Instead, the user must actively seek out the data by clicking the “add data feed” button in the lower right and then choosing from a list of available feeds. The hope is that experiencing the need for data, and being able to successfully act on it, in the Sim will help the player to feel a need for action in the real world.

The rules of the Sim require the player to finish the project on time. The first problem a user must recognize and solve is that productivity and quality (i.e. the fraction of work done being done correctly) are less than explicitly planned. Consequently, if the originally planned staffing levels are maintained, the project will be unfinished at the originally scheduled completion date and the Sim player won’t succeed.

Figure 2: Productivity, Staff, and Quality. Blue “skinny” arrows represent causal connections (as opposed to physical flows, represented by double arrows). Mathematically, correctlyDoingWork= qualityt * doingWorkt. And, incorrectlyDoingWork= (1 -qualityt) * doingWork. If staffing is maintained as planned while productivity and quality are lower than planned, less work will get done and more of it will need to be redone than planned. At any point in time, less of the project will be complete than planned.

In order to complete the simulated project within the allotted time, the player must use his reserves to boost staffing above the planned levels. Doing so isn’t difficult: the Sim player simply must move the Staffing Lever (at the bottom of the main window) above 100% of plan. But, how does the player know how much to increase staffing, indeed how does he know that he must increase staffing at all?

“Expenditures”, the information available on the Sim’s main window, does not reveal the problem. Expenditures data relates to how many people are working, not how much they are accomplishing (see Figure 3). To recognize the need for, and to decide how much, additional staffing is necessary, the player needs a deeper view into the project than is provided by the main window’s information on expenditure.

Figure 3: Expenditures. Expenditures (spend) is connected to (or caused by) staff, not by how much work has actually been done. Staff levels can be deduced from data on spend. But, if productivity and quality are unknown, spend data does not provide enough information to deduce whether the project is on track or not.

The Progress Data Feed provides the deeper view that the player needs. The data feed is not perfect; it represents what could be available to a project manager relatively easily. Because Undiscovered Rework is not directly measurable, the data feed does what most people do – simply counts the work that has been “done” and doesn’t worry about the portion that will need to be redone. That is, the progress data feed shows perceived cumulative work done, which is the sum of Correctly Done Work and Undiscovered Rework. Imperfect though it is, the progress data feed (and a bit of hand-eye coordination) allows the player to use the staffing lever to keep the project’s cumulative (perceived) progress on track with planned progress (see Figure 4).

Figure 4: The player moves the slider, to make perceived cumulative work done track the plan (i.e. the “budget”).

The progress data feed allows the player to create — indeed to be part of — a critical control loop in the simulated project: When perceived cumulative work done drops below plan, the player increases the staffing slider and that action brings cumulative work done back up.

Figure 5: A critical control loop. The Progress Data Feed shows work perceived done and planned work done. The player mentally considers the amount of work done relative to planned and adjusts staff accordingly. 

The game designers wanted the player to make a conscious effort to open a data feed, to avoid a situation where players would open data feeds simply because they had nothing else to do. This led to three additional design decisions. First, opening a data feed costs “money”, which is subtracted from reserves. Second, opening a data feed is a bit time consuming (the user sees three different windows, and must click four times in the process of opening a data feed). Third, and most importantly, the game player has a “distraction” clamoring for her attention.

The “distraction” in the Sim involves a realistic complication: In order to succeed, a Sim player must not only complete the LV project within a specified time, the completed LV must be able to lift the expected mass of the crew vehicle. The problem is that the expected mass of the crew vehicle continues to increase as the LV project continues. In response to these increases, the Sim player must initiate design changes to compensate. Virtually the entire left side of the Sim’s main window is devoted to the issue of lift and changes in the LV design. The intent was to make this aspect very obvious and, thus, very distracting.

Interviews with users indicate that the Sim can accomplish this fundamental purpose among those who haven’t already gained this insight through real-life experience.

For complete details about the learning simulation and the results observed from using it, please read our white paper, available for download below.