In this paper a description is given of an application of DSTO's BattleModel [1] using agents and... more In this paper a description is given of an application of DSTO's BattleModel [1] using agents and humansin-the-loop. This application is being developed in response to requirements from the Royal Australian Air Force (RAAF). In preparation for upgrades to the avionics, electronics and sensors of the P-3C Orion fleet, the RAAF is developing new operational procedures and tactics. To assist in the test and evaluation process, operational analysis is being conducted in the Air Operations Division (AOD, DSTO). The aim of this paper is to describe the development and use of an analysis capability used by AOD that is unique in operating in both constructive and Crew-in-the-Loop simulation modes. The defining difference between these two modes is 'who' (or 'what') makes the tactical decisions during mission operation. In the case of constructive simulations, tactical decision making is by BDI (Belief-Desire-Intention) agents that have been programmed using the dMARS (distributed Multi-Agent Reasoning System; [2, 3]) language. In the case of Crew-in-the-Loop simulations, decision-making is by actual human crewmembers who replace agents, and interact with the simulation environment via graphical user-interfaces (GUIs). BattleModel is the common software architecture and run-time environment for both forms of simulation. It is a framework that enables definition, integration and deployment of various models and applications in fully digital spaces, or in spaces where the digital and real-worlds act in concert. It thus supports the 'plugging-in' of humans and agents into the same simulation world. The design and construction reported here is significant because the crew-in-the-loop simulation capability has involved substantial re-use of the architecture and models that were also used for the initially developed constructive simulations. Humans and agents are interchangeable as a consequence of agent-oriented design.
Within Air Operations Division of DSTO 1 intelligent agents are used to model the tactical decisi... more Within Air Operations Division of DSTO 1 intelligent agents are used to model the tactical decision making processes of pilots and ghter-controllers involved in air combat. One of the largest hurdles to be overcome by software engineers and analysts, when developing simulations of the air defence environment, is the acquisition of domain knowledge. Primarily the source of this knowledge is the pilots and other operational personnel, whose availability is limited and who have little experience with the design or development of simulation software. The adoption of agent oriented technologies has realized a number of signicant bene ts. High amongst these is the ability for operational air force personnel to become actively involved in the modi cation, design and development of these simulations. This involvement has dramatically reduced the time taken to prototype, test, and commission software and has resulted in simulations that have the con dence of the RAAF.
Whitestein Series in Software Agent Technologies and Autonomic Computing, 2008
Since 1990 a focused intelligent agent research and development programme within the Defence Scie... more Since 1990 a focused intelligent agent research and development programme within the Defence Science and Technology Organisation (DSTO) has underpinned a strong history of deployed operational simulations. Originally aimed at improving simulations of fighter ...
In this paper a description is given of an application of DSTO's BattleModel [1] using agents and... more In this paper a description is given of an application of DSTO's BattleModel [1] using agents and humansin-the-loop. This application is being developed in response to requirements from the Royal Australian Air Force (RAAF). In preparation for upgrades to the avionics, electronics and sensors of the P-3C Orion fleet, the RAAF is developing new operational procedures and tactics. To assist in the test and evaluation process, operational analysis is being conducted in the Air Operations Division (AOD, DSTO). The aim of this paper is to describe the development and use of an analysis capability used by AOD that is unique in operating in both constructive and Crew-in-the-Loop simulation modes. The defining difference between these two modes is 'who' (or 'what') makes the tactical decisions during mission operation. In the case of constructive simulations, tactical decision making is by BDI (Belief-Desire-Intention) agents that have been programmed using the dMARS (distributed Multi-Agent Reasoning System; [2, 3]) language. In the case of Crew-in-the-Loop simulations, decision-making is by actual human crewmembers who replace agents, and interact with the simulation environment via graphical user-interfaces (GUIs). BattleModel is the common software architecture and run-time environment for both forms of simulation. It is a framework that enables definition, integration and deployment of various models and applications in fully digital spaces, or in spaces where the digital and real-worlds act in concert. It thus supports the 'plugging-in' of humans and agents into the same simulation world. The design and construction reported here is significant because the crew-in-the-loop simulation capability has involved substantial re-use of the architecture and models that were also used for the initially developed constructive simulations. Humans and agents are interchangeable as a consequence of agent-oriented design.
Within Air Operations Division of DSTO 1 intelligent agents are used to model the tactical decisi... more Within Air Operations Division of DSTO 1 intelligent agents are used to model the tactical decision making processes of pilots and ghter-controllers involved in air combat. One of the largest hurdles to be overcome by software engineers and analysts, when developing simulations of the air defence environment, is the acquisition of domain knowledge. Primarily the source of this knowledge is the pilots and other operational personnel, whose availability is limited and who have little experience with the design or development of simulation software. The adoption of agent oriented technologies has realized a number of signicant bene ts. High amongst these is the ability for operational air force personnel to become actively involved in the modi cation, design and development of these simulations. This involvement has dramatically reduced the time taken to prototype, test, and commission software and has resulted in simulations that have the con dence of the RAAF.
Whitestein Series in Software Agent Technologies and Autonomic Computing, 2008
Since 1990 a focused intelligent agent research and development programme within the Defence Scie... more Since 1990 a focused intelligent agent research and development programme within the Defence Science and Technology Organisation (DSTO) has underpinned a strong history of deployed operational simulations. Originally aimed at improving simulations of fighter ...
Uploads
Papers by Martin Cross