Hbm Depot

Period of Performance: 09/11/2008 - 03/11/2009

$80K

Phase 1 SBIR

Recipient Firm

Soar Technology, Inc.
3600 Green Court Array
Ann Arbor, MI 48105
Principal Investigator

Abstract

Current military M&S systems for training and experimentation rely on encoded human behavior models (HBMs) that must be produced and maintained by teams of subject matter experts (SMEs) and professional knowledge engineers (KEs) in an arrangement that can be time-consuming and expensive. The programming by demonstration (PBD) and machine learning paradigms offer solutions, but are not without challenges, typically producing behaviors that are too simple for operational use, require extensive supplementation by a KE and/or require large sets of examples. Soar Technology has formed a team rich in experience with Navy modeling and simulation, HBM development, artificial intelligence, and serious games for training to propose HBM DEPOT, a tool that supplements simulation-based demonstrations with a novel and intuitive guided after-action whiteboarding session based on existing research and tools. This will result in a system that captures demonstrations of Navy Aviation tactics using a low-cost Delta 3D Navy flight simulator and supplements them with an easy to use diagrammatic representation (pre-loaded with Navy Aviation general domain knowledge) and learning algorithm to generate high-quality HBMs. The HBMs will be suitable for use in any virtual environment where intelligent computer generated forces (CGFs) or non-player characters (NPCs) such as a wingman or sophisticated OPFOR are required.