Measuring the complexity of simulated engineering design problems

Abstract

The ability of tomorrow's engineering professionals to solve complex real-world problems is dependent on their education and training. We posit that engineering education and training in design would be improved by presenting students with design challenges with increasing levels of complexity as they advance in engineering curricula. In order to construct design challenges with increasing levels of complexity, a framework for assessing the complexity of engineering design problems must be developed. As a first step toward this goal, we consider the complexity of simulated design problems, which have been previously developed as part of virtual engineering internships and which have the advantages of being well-defined and solvable. In this paper, we present a parameterized, mathematical model to quantify engineering design problem complexity. In particular, we present three functions that model the process by which a student moves from information provided and assumptions to predicting design performance and then to a final design choice. These functions are F, students' predictions of device performance, V, how students value performance criteria, and P how students develop preferences for specific designs. Finally, based on this framework for quantifying simulated design problem complexity, we present a metric of complexity, tractability T, supported by data from real student work on a simulated engineering design problem.