Technology evaluation is a resource evaluation problem compounded by uncertain future budgets and the unknown potential of new emerging technologies. In contrast to traditional resource allocation approaches that are usually based solely on expert judgment or subjective opinions, this presentation will demonstrate how physics-based design methods can be used to quantify technology potential and bound uncertainty in the systems engineering process. In recent years, systems engineering has become increasingly complicated by the focus on integrated “systems-of-systems” as opposed to unitary stove-piped systems. This integration challenge extends the physics-based design paradigm to a higher level of abstraction and requires a larger “world-view” of the problem. While such analysis is technically feasible with today’s computerized design tools and growing computational power, the myriad of data generated cannot be readily turned into actionable information for decision makers. Design methodologies developed the Georgia Institute of Technology’s Aerospace Systems Design Laboratory (ASDL) begin to address this challenge by infusing rapid domain-spanning design exploration methods and advanced visualization capabilities. When techniques such as design-of-experiments, modeling and simulation, surrogate modeling, neural network models, and probabilistic techniques are coupled with advanced multi-dimensional visualization capabilities, “visual analytics” become possible for the first time. An Integrated Product Team of subject matter experts can exercise a simulation in a collaborative environment in near real-time using the concept of surrogate models: highly accurate approximations of existing high-fidelity tools. Furthermore, once surrogates are created around a contextualized problem, designs can be explored in a fraction of the time as traditional methods. Using this ability as an enabler, this presentation will address how probabilistic techniques such as Monte Carlo Simulation can be used not only for uncertainty quantification, but also to flood the design space with a myriad of possible solutions and parametrically filter the space to reveal solutions that meet customer needs and maximize overall effectiveness. Next, these techniques are extended to provide users with the ability to “design-for-capability.” This innovative approach uses a technique matured at ASDL called inverse design to look at the problem using a top-down decomposition of capability-based metrics as opposed to a bottom-up, system-centric view of the problem. Potential solution architectures that maximize effectiveness can be explored and queried in a graphical environment and decision makers can play “what-if” games on-the-fly. In addition to the development of the aforementioned methodologies, ASDL has also prototyped a Collaborative Visualization Environment (CoVE) that leverages government investments in large-format visualization technologies to facilitate decision making across large dimensionality problems. As such collaborative environments become more affordable for industry and government partners, the notion of multi-site, collaborative systems-of-systems design is possible. An example of a collaborative approach to the evaluation of emerging technologies will be demonstrated and both examples from the civil and military worlds will be shown.
