This paper lays out a methodology for evaluating system-level requirements feasibility during System-of-Systems (SoS) conceptual design, especially in the absence of historical data, and simultaneously documents the methodology's application to the design of small aerial vehicles. Developed as part of a larger effort to set requirements for autonomous robots operating to support combat troops in urban environments or otherwise complex terrain, the methodology manifests itself in a parametric, interactive, and virtual design environment. The proposed approach relies on a comprehensive set of parametric models to estimate subsystem performance characteristics from user-supplied subsystem design parameters. These models are then integrated such that, through optimization and the use of response surface methodology, the system-level performance characteristics can be rapidly computed. This allows the environment's user to visualize design tradeoffs across the entire design space, quantify the sensitivity of system-level performance measures to subsystem design parameters, and evaluate the feasibility of system-level requirements. It will be shown that, in the current application, this methodology allows the designer to identify the feasible portion of the system-level design space and thereby dramatically narrow the scope of future SoS design efforts. IMADE (Integrated MAST Analysis and Design Environment) is a GUIbased implementation of the methodology as it was applied to the design of small aerial vehicles. Example analyses from IMADE demonstrate that a quadrotor designed for a hypothetical 10-minute reconnaissance mission will need to be approximately a third of a meter wide but could shrink to a quarter of a meter wide if only the lidar were excluded. These analyses also show that the quadrotor is capable of transmitting video for hours if perched in position and will be visible from hundreds of meters away.

No abstract