A novel navigational planning scheme for autonomous underwater vehicles

A navigational planning system for autonomous under-water vehicles (AUVs) that employs case-based reasoning (CBR) techniques is described in this paper. CBR is a reasoning technique using past episodic information and offers several advantages such as faster synthesis of plans, incremental learning, and easy knowledge engineering. We discuss the design of an annotated map system for representing the navigational environment, a case-based path planner that utilizes old routes to generate new ones, and an obstacle avoidance scheme. Navigational planner described is part of a larger integrated mission planning system for AUVs.<
>