Multi-Attribute Tradespace Exploration and its Application to Evolutionary Acquisition

The Air Force has recently embraced Evolutionary Acquisition (EA) as its acquisition strategy of choice. EA is an especially difficult method of acquisition and presents some extraordinary challenges at the system engineering level.
Multi-Attribute Tradespace Exploration (MATE) is a tool at the system engineer’s toolbox that can provide some focus on a project in EA. MATE, a tool initially developed by Adam Ross and Nathan Diller at MIT, is a method of developing models to simulate the product user’s preferences for the attributes of a design. Once these preferences are well known, they can be used to guide the design choice.

The design choice is further guided by the creation of system level computer models that represent the design choices available to the engineer. These choices are then varied systematically to create a “tradespace” of possible designs. This tradespace exhaustively enumerates all of the possible design choices for the engineer. Then, through the preference models previously developed, each possible design is ranked in order of user utility and cost. The result can be graphed, giving a visual representation of the utility and cost of literally thousands of architectures in a single glance.

It might be possible to create multiple models for a system that anticipates the different evolutions that might be implemented on it. Through such models, intuition about the initial system configuration might be enhanced, and the initial architecture choice modified based on the information gained from this modeling.