Rationalizing source data: tailoring geospatial data for advanced distributed simulation

ABSTRACT
The production of geospatial environments for modeling, simulation, and training starts with the acquisition of suitable source data, imagery, digital elevation models, various vector data layers, and representative or geospecific 3D models. Whether the resulting modeling and simulation runtimes are fully compiled or created "on the fly" for presentation, significant effort must be expended to rationalize disparate data sources prior to the database production. In this paper, we describe a set of source data preparation tools that have been developed to address the generation of derived data sets. These tools are tailored to the requirements of the database generation process and have been adapted to the production of datasets which support correlated runtimes across visual, constructive, and serious game runtimes.

Specifically, we will discuss issues in the preparation of large area digital elevation models (DEMs) and the generation of road networks and surface material maps from remotely sensed imagery. We will focus on the unique requirements of advanced ground simulation, such as including expanded environmental data models that contain detailed attribution and tailoring databases to address differences and limitations inherent across multiple simulation runtimes. These issues are particularly important in networked simulations where different simulations are running in an overall distributed exercise and are expected to be correlated in terms of mobility, sensor simulation, cultural content, and intrinsic properties of the terrain environment.

Specific examples will be taken from a range of constructive (OneSAF, JointSAF, JCATS) and serious game (Steel Beasts, VBS2, Havok) runtimes.

VITA
Mr. David M. McKeown has been involved in environmental issues for modeling and simulation since the early '90s, beginning with his involvement with the DARPA SIMNET program. He has actively participated in technical conferences for IEEE, IMAGE Society, etc. and gave a well-received tutorial on environmental considerations for pattern of life (POL) at GameTech 2012. He is President of TerraSim, Inc and an Adjunct Research Professor in Computer Science at Carnegie Mellon University in Pittsburgh.