A Situation Assessment System for the Msel Eave-III Auvs

Abstract

This paper describes a situation assessment system for an autonomous underwater vehicle (AUV). This system is a result of research on knowledge-based guidance systems applied to AUVs, which is currently taking place at the Marine system Engineering Laboratory, University of New Hampshire. This paper gives the reader some background on the project, an introduction to the MSEL KBS architecture, and a description of how the situation assessment system is implemented. In a dynamic world with a complex mission, a mobile robot needs the ability to analyze its environment and its situation. We call this situation assessment. Since unplanned situations will most likely occur, the robot will need some ability to perform qualitative reasoning on its situation, before it modifies its original plan. This qualitative reasoning implies a system capable of building and maintaining a model of the vehicle's environment and itself. It also involves use of data from several different sources (sensor fusion) and combining these data into more abstract terms.