Managing complexity of aerospace systems

Shashank Tamaskar, Purdue University

Abstract

Growing complexity of modern aerospace systems has exposed the limits of conventional systems engineering tools and challenged our ability to design them in a timely and cost effective manner. According to the US Government Accountability Office (GAO), in 2009 nearly half of the defense acquisition programs are expecting 25% or more increase in unit acquisition cost. Increase in technical complexity has been identified as one of the primary drivers behind cost-schedule overruns. Thus to assure the affordability of future aerospace systems, it is increasingly important to develop tools and capabilities for managing their complexity. We propose an approach for managing the complexity of aerospace systems to address this pertinent problem. To this end, we develop a measure that improves upon the state-of-the-art metrics and incorporates key aspects of system complexity. We address the problem of system decomposition by presenting an algorithm for module identification that generates modules to minimize integration complexity. We demonstrate the framework on diverse spacecraft and show the impact of design decisions on integration cost. The measure and the algorithm together help the designer track and manage complexity in different phases of system design. We next investigate how complexity can be used as a decision metric in the model-based design (MBD) paradigm. We propose a framework for complexity enabled design space exploration that introduces the idea of using complexity as a non-traditional design objective. We also incorporate complexity with the component based design paradigm (a sub-field of MBD) and demonstrate it on several case studies. The approach for managing complexity is a small but significant contribution to the vast field of complexity management. We envision our approach being used in concert with a suite of complexity metrics to provide an ability to measure and track complexity through different stages of design and development. This will not only lead to simpler designs but also help the designers calculate the impact of their design decisions on integration cost.

Degree

Ph.D.

Advisors

DeLaurentis, Purdue University.

Subject Area

Aerospace engineering

COinS