Using Dissimilarity Metrics to Identify Interesting Designs
NASA’s Jet Propulsion Laboratory, Pasadena, California
A computer program helps to blend the power of automated-search software, which is able to generate large numbers of design solutions, with the insight of expert designers, who are able to identify preferred designs but do not have time to examine all the solutions. From among the many automated solutions to a given design problem, the program selects a smaller number of solutions that are worthy of scrutiny by the experts in the sense that they are sufficiently dissimilar from each other. The program makes the selection in an interactive process that involves a sequence of datamining steps interspersed with visual displays of results of these steps to the experts. At crucial points between steps, the experts provide directives to guide the process. The program uses heuristic search techniques to identify nearly optimal design solutions and uses dissimilarity metrics defined by the experts to characterize the degree to which solutions are interestingly different. The search, data-mining, and visualization features of the program were derived from previously developed risk-management software used to support a risk-centric design methodology.
This program was written by Martin Feather of Caltech and James Kiper of Miami University, Oxford, Ohio, for NASA’s Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free on-line at www.techbriefs.com/tsp under the Software category.
This software is available for commercial licensing. Please contact Karina Edmonds of the California Institute of Technology at (626) 395-2322. Refer to NPO-40456.
This Brief includes a Technical Support Package (TSP).
Using Dissimilarity Metrics To Identify Interesting Designs (reference NPO-40456) is currently available for download from the TSP library.
This week's Question: This fall, Starship Technologies, an Estonia-based startup created by two Skype co-founders, will begin testing its autonomous delivery robot in Washington, D.C. Washington is the first U.S. municipality to approve...
Subscribe today to receive the INSIDER, a FREE e-mail newsletter from NASA Tech Briefs featuring exclusive previews of upcoming articles, late breaking NASA and industry news, hot products and design ideas, links to online resources, and much more.