An Overview of Agent Technology for Satellite Autonomy

May 1999

Proceedings of the Twelfth International Florida Artificial Intelligence Research Society Conference (FLAIRS 1999)

Volume

Proceedings of the Twelfth International Florida Artificial Intelligence Research Society Conference (FLAIRS 1999)

All Papers

To operate the majority of today’s satellites is costly and manpower intensive. In order to reduce cost, enhance survivability, and increase responsiveness, technologies are needed which will make satellites more autonomous. Many artificial intelligence (AI) based prototypes have been developed with the above objectives in mind, however these prototypes are typically geared towards solving a particular problem with a particular technique such as identifying faults in the attitude control system using a model-based reasoning system. Generally these AI systems only look at a subset of the total information available and do not take a system approach to reasoning. To optimize the ability of an AI system to make correct decisions many factors need to be considered including the status of all subsystems, current and future mission objectives, and the status of components external to the satellite including ground facilities and other satellites. Intelligent agent technology provides a mechanism to integrate these various components and to reason at a higher level. The purpose of this paper is to give an overview of intelligent agents as they apply to space systems. The paper begins with a short introduction of what intelligent agents are and how they operate. Following this we will highlight particular areas where the use of intelligent agents may be advantageous. Next we highlight some of the intelligent agent efforts which are ongoing in the Space Vehicles (VS) Directorate of the Air Force Research Laboratory (AFRL). Following this we will present an overview of some of the ongoing efforts in the space community external to AFRL. We will then present an overview of some of the agent-based architectures which are publicly available and discuss their applicability to the space domain. Lastly we will summarize and discuss future directions in this area.