A modular design tool for visualizing complex multiscale systems
Editor: Udo Lindemann, Srinivasan V, Yong Se Kim, Sang Won Lee, John Clarkson, Gaetano Cascini
Author: Egan, Paul; Cagan, Jonathan; Schunn, Christian; LeDuc, Philip
Institution: 1: Dept. of Mechanical Engineering, Carnegie Mellon University, United States of America; 2: Dept. of Psychology, University of Pittsburgh, United States of America
As engineers design ever more complex systems, such as bionanotechnology, it is particularly difficult for engineers to navigate the influences of parameters to design across multiple scales. We introduce a software tool that contains graphical interface modules that display multiple inputs and outputs simultaneously in real-time, across multiple scales of interest. To assess the utility of real-time feedback in the tool, the tool is validated through user studies involving counterbalanced sequences of optimization tasks with varying complexity and either real-time or delayed feedback. Real-time feedback consistently improved a userâs ability to find better designs and the benefit increased as task complexity grew. Our scientific contribution is a greater understanding of how users search and understand complex design spaces. Users in the real-time condition were made fewer changes to designs each iteration and identifying such strategies could inform agent-based algorithms. Findings aid complex systems applications including virtual experiments, reverse engineering, and scientific discovery.