A peer-reviewed paper on “Tool Support for Parametric Analysis of Large Software Simulation Systems” was accepted for the Tools-demonstration session of the 2008 Automated Software Engineering Conference (ASE 2008), to be held in September 15-18 2008 in L’Aquila, Italy. ASE is the major conference in the field of Automated Software engineering and the flagship conference for the Robust Software Engineering (RSE) group at NASA Ames. During this event, RSE researchers will demonstrate a tool which has been developed by Ames, the Jet Propulsion Laboratory (JPL), and West Virginia University. It intelligently generates simulation parameter variations using a combination of Monte Carlo, combinatorial n-factor generation, and model-based test case generation. Data produced by simulation runs are analyzed using multivariate clustering algorithms (generated by the Ames AutoBayes tool) and the machine-learning tool TAR3, which facilitates root-cause analysis.
This tool is being used for the analysis of abort and re-entry scenarios for the Orion spacecraft, and will be demonstrated on a computational model of the Ames Hover Test Vehicle (HTV).
PROJECT TEAM: Johann Schumann, RIACS/NASA Ames; Karen Gundy-Burlet, NASA Ames; Corina Pasareanu, PSGS/NASA Ames; Tim Menzies, West Virginia University; Tony Barrett, JPL
NASA PROGRAM FUNDING: ISD
Contact: Johann Schumann