An AI Agent Construct for Space and Planetary Science Applications

Many of today's artificial intelligence methods are simply another type of algorithm or model. The key to real intelligent systems lies in making the systems adaptive. It is toward this goal that we must work, in cooperation with technical staff from all affected disciplines. This paper stresses the introduction of adaptation in intelligent systems for space or planetary science systems applications. Adaptation is important to the long-term use of intelligent systems. In this direction, a practical framework, or construct, is suggested for the construction of an intelligent Agent for use in space or planetary science systems. The construct proposed is based upon integrated/hybrid artificial intelligence systems (neural networks, expert and procedural systems combined with intelligent, microelectronic sensors and hardware).