Skip navigation
PDF 966 KB
DOI: http://dx.doi.org/10.7551/978-0-262-33027-5-ch098
Pages 562–569
First published 20 July 2015

Towards a methodology for describing the relationship between simulation and reality

Eric Schneider, Elizabeth I. Sklar, M. Q. Azhar, Simon Parsons, and Karl Tuyls

Abstract

For research that carries out experiments in simulation, an important question is how the results will translate into the real world. This paper proposes a method for comparing results obtained in simulated versus physical environments, based on interval relationships between metrics gathered in both settings. The approach is motivated by the fact that the relationship between absolute measures often does not tell much. For example, the amount of time taken to complete a task in simulation versus the same task in the physical world could always be shorter in simulation because of speed-up factors embedded in the simulator. Three different metrics are introduced that describe different interval relations, and these are demonstrated using two case studies.