Air Operations Enterprise Modelling

Project funded under the DARPA JFACC program

AIAI Staff: John Kingston, Jessica Chen-Burger, Lea Ruscio

AIAI worked on the “Agile Control of Military Operations” project as part of the BBN Technologies Air Operations Enterprise Modelling team. AIAI’s original aims of carrying out model integration and development, requirements capture & experiment analysis, model configuration & modification and model assessment were modified to meet the revised requirements of the project as the project progressed. AIAI therefore carried out three major work items:

• Multi-perspective modelling. This provided the theoretical basis for our analysis of enterprise modelling, and in particular for the selection of modelling techniques to represent the results of knowledge acquisition. We examined issues such as ontology, appropriate modelling techniques, and automatic consistency checking.
• Knowledge acquisition. We initially acquired some knowledge from existing IDEF0 models as part of our multi-perspective modelling work, and represented this knowledge using Business System Development models, working on the basis that an air operations enterprise could be represented using similar modelling techniques to other business enterprises. However, it was felt that a more coherent, complete and consistent set of models could be obtained by carrying out knowledge acquisition with a recognised expert on air operations. It was also felt there would be benefits in using a more widely known modelling technique. This knowledge acquisition was therefore carried out, and the results modelled using a combination of IDEF0, Role Activity Communication Diagrams, and UML.
• Intelligent Adversary. The Intelligent Adversary is a program that combines two AI techniques – rule based programming and genetic algorithms – to produce a planning system that directs military units to attack and/or defend key targets. It is an “adversary” because it is designed to play the Red side against various controllers based on control theory that direct the Blue units; it is “intelligent” because it incorporates some principles of military strategy and tactics whilst trying to optimise the use of available military resources. The Intelligent Adversary has been developed to a stage where it can be considered to be an operational prototype.

Publications:

Knowledge-based Automatic Verification & Validation for Business Models. Y-H Chen-Burger, D. Robertson and J. Stader. Proceedings of the 1^st Workshop on Advances in Enterprise Control

A Knowledge Based Multi-Perspective Framework For Enterprise Modelling. Y-H Chen-Burger. AIAI Technical report, 2001.

Knowledge Sharing and Inconsistency checking on Multiple Enterprise Models. Y-H Chen-Burger. Proceedings of the Knowledge Management & Organizational Memory Workshop, IJCAI-2001, Seattle, Washington, 6 August 2001.

Knowledge Models in BSDM.Y-H Chen-Burger. AIAI Project report.

Knowledge Models in RACD and UML. Y-H Chen-Burger and L. Tonnesson. AIAI Project report.

Ruscio, L., Levine, J. and Kingston, J. (2000). Applying Genetic Algorithms to Hierarchical Task Network Planning. Proceedings of the19th Workshop of the UK Planning and Scheduling Special Interest Group (PLANSIG 2000), Open University, Milton Keynes, UK, December 2000.

The Intelligent Adversary: Combining Knowledge Based Systems and Genetic Algorithms for Military Planning. AIAI project report. Will be submitted to an appropriate journal or conference.