June 26 - 30, 2004
Saturday to Wednesday
Seattle, Washington, USA

 

 

Session:

MSA - Military and Security Applications of Evolutionary Computation

Title:

Genetic algorithms for UGV navigation, sniper fire localization, and unit of action fuel distribution

   

Authors:

Talib Hussain
David Montana
Marshall Brinn
Daniel Cerys

   

Abstract:

BBN Technologies is applying genetic algorithms (GAs) to create state-of-the-art solutions in three different military applications. For the navigation of multiple unmanned ground vehicles, we apply a GA to determine vehicle paths that accomplish a set of mission goals. The GA incorporates domain-specific operators that make changes based on good tactics, and it evaluates the quality of the paths against multiple tactical criteria. In the Boomerang project, we apply a GA to identify the source of sniper fire in real-time. The system is currently being deployed in Iraq on military HMMVVs. To distribute fuel to the combat vehicles within a Future Combat Systems Unit of Action (UA), we apply a GA as an agent within a larger agent system that forecasts fuel demand over a 72-hour period. The GA agent schedules a fleet of refueling trucks to deliver fuel directly to the combat vehicles at the times and locations requested.

Home

Program

Search

Author Index

Sponsors

Committee

Contact Us

Help