Enhanced Situation Awareness using Random Particles
Abstract
Modern command and control systems present the current view of the situation through a situation picture that is being built up from fused sensor data. However, merely presenting a comprehensible description of the situation does not give the commander complete awareness of the development of a situation. This article presents a generic tool for prediction of forthcoming troop movements. The technique is similar to particle filtering, a method used for approximate inference in dynamic Bayesian networks. The prediction tool has been implemented and installed into an existent electronic warfare system. The tool makes use of the system's geographic information system to extract geographic properties and calculate troop velocities in the terrain which is, in turn, being used for the construction of the tool's transition model. Finally, the result is presented together with the situation picture. The prediction tool has been evaluated in field tests performed in cooperation with the Swedish Armed Forces in an exercise in Sweden during the spring of 2005. Officers and operators of the electronic warfare system were interviewed and exposed to the tool. Reactions were positive and prediction of future troop movements was considered to be interesting for short-term tactical command and control.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2005
- Accession Number
- ADA463850
Entities
People
- Joel Brynielsson
- Jonas Nordh
- Lennart Voigt
- Mattias Engblom
- Robert Franzen
Organizations
- Royal Institute of Technology