The Modeling and Simulation industry has long been plagued by geospatial database representational flaws and miscorrelation used to represent the synthetic natural environment within military training systems. These errors spawn from a wide range of sources, including design decisions, performance simplifications, bad source data, and unrealistic or erroneous database content. These errors are so pervasive across systems that they are often accepted as inevitable despite Soldier training impacts.

This paper discusses the work conducted under a Phase II SBIR. It provides proposed solutions consisting of toolsets that assess and compare geospatial and geometric data between disparate database formats and representations while providing multiple testing mechanisms such as visual inspection, automated testing and interactive testing using reusable software libraries and analysis artifacts. Real world examples of specific database errors on Army simulation programs will illustrate the complexity of tying geometric flaws with training impact.

This paper examines the challenges, planned approaches, and solutions for both detection and evaluation of correlation and representation errors. The work includes implementation of a testing framework and open standards for test tools and test data exchange, as well as instantiation of that framework in the C-nergy toolset. Moreover, technical transfer of this research by leveraging emerging common Army standards, such as SE Core and One Semi Automated Forces (OneSAF), is critical to successful widespread use of correlation testing toolsets.