In our previous Common Elements blog , we gave an overview of the new Common Element models, showing how they’re structured to provide better visibility and control in building your cost estimate. Each of these new models also offers multiple methods of estimating, from models informed by information from the Prime Mission Product estimate, to standalone models, to simple rule-of-thumb estimating using cost factors. We’ve done a great deal of new research to build these different types of models, and many of the new common elements offer all 3 methods.
The “Informed by Prime Mission Product (PMP)” approach bases costs on the inputs for the common element model object but is also informed by other PMP cost object inputs and results. We use this approach for the new System Test and Evaluation (T&E) model, which estimates cost based on PMP estimate results for the non-recurring development cost, and specific T&E model inputs, like number of test launches (for missile T&E), Design Complexity, and Test Requirement Complexity. This methodology is primarily based on a RAND study into T&E costs, but we’ve expanded on this by incorporating additional data (internal and from other publicly-available studies), expanding the scope to other system types, and adding additional cost drivers to refine the results. For example, additional cost drivers like Design Complexity (limited enhancement vs. new state-of-the-art design) and Test Requirement Complexity (testing by multiple services with their own unique requirements, compressed schedules adjustment, etc.) can be used to fine-tune the estimate.
The “Standalone” estimating methodology applies a separate model informed only by the specific inputs asked by the model. For example, the Training common element model estimates courseware development and delivery, and asks questions about the content (How many hours of instruction? How complex is the content? How will it be delivered? How many people will receive training? Etc.) The standalone approach is needed, as Training needs vary greatly from program-to-program and does not generally correlate well with any information from the PMP estimate. Another example is the Systems Engineering common element, which can be estimated by the Standalone COSYSMO cost model, developed at USC. The model, its associated inputs and resulting outputs are wholly included in the Systems Engineering model architecture.
Finally, nearly all common element models allow for simple cost estimation using basic Cost (or Labor Hour) Factors. Some organizations have defined “rules of thumb” for estimating below-the-line costs; for example: Systems Engineering costs equal 20% of estimated Development costs for all the subsystems and components.
By providing many different estimating approaches, it improves flexibility and makes it easier to deliver new research. For more information on all 12 new models, their structure and supporting research, please sign up for our Common Elements Webinar.. The common element models will debut soon with the upcoming release of TruePlanning 16.2.