Trade-off Analytics

Chapter 3Quantifying Uncertainty

Robert F. Bordley

Integrated Systems & Design, University of Michigan, Ann Arbor, MI, USA; Defense & Intelligence Group, Booz Allen Hamilton, Troy, MI, USA

 

Laplace: The theory of probability is at bottom nothing but commonsense reduced to calculus.

3.1 Sources of Uncertainty in Systems Engineering

Best practice in systems engineering identifies all relevant factors outside the control of the design team. Some of these factors, for example, certain external interfaces, are completely known to the team. But “many phenomena that are treated as deterministic or certain are, in fact, uncertain,” (OMB guidance, circular 94, 2015). The inappropriate treatment of uncertain quantities as deterministic, also called the flaw of averages (Savage et al., 2012), can lead to design decisions that could prove very costly if the actual value of this quantity is very different from the average. So the team must acknowledge their uncertainty about both

1. what is feasible and affordable, for example, technological, integration, interface, and supply uncertainties and
2. what is desirable, for example, uncertainties about stakeholder preferences, adversary reactions, and environmental changes impacting stakeholders and adversaries.

There is added intrateam uncertainty due to different specialists within the design team not knowing how they will be impacted by the choices and discoveries of other specialists.

In addition to acknowledging these uncertainties, the team ...