We have three key components to our research.
First, we are building a resource that allows systems-dynamics models (and other model types) to interoperate in a federation fashion, allowing the ad hoc creation of “super-models” that can run simultaneously, as if the resulting supermodel was created as a single model. At the same time, we enable the reluctant stakeholders to have their confidential models and data interoperate with models and data from other stakeholders (with perhaps conflicting interests), allowing the joint resources to be applied to a common problem without exposing sensitive data or algorithms between stakeholders. The resource we are building will also allow different model types, for instance system dynamic models and agent-based models, to interact.
Second, we are building resources to visualize the findings of these interacting models. Key to this is enabling the non-scientist decision-maker to understand the implication of the results of exercising the “super-models” (above), particularly the sensitivities of those models to various inaccuracies and unknowns. This new, implicit visualization is critical to our third key area.
Finally, our third key area of our research is applies three-dimensional Bayesian analysis techniques to quickly identify where additional research is necessary before a non-scientist manager can make an informed decision. The goal is rapid decision analysis of complex and mission-critical issues.