(:notitlegroup:)
A new behavior, called BHV_InformationGain has been developed to plan paths to efficiently cover an area of seabed with an acoustic sensor, such as a sidescan sonar, to detect mines-like objects. This behavior can be used in place of pre-specified waypoints. The path generated is more efficient than the classical lawn mower path, particularly in complex or non-convex environments. The performance of a sidescan sonar as a function of range is highly non-linear and is a affected by environmental factors. As a result, it is difficult to find a closed form for the Actual Objective Function (AOF). The Reflector tool provided with MOOS-IvP is used to sample the underlying objective function by simulating moves over a discretized subset of next possible headings, in real-time, to determine the Expected Entropy Reduction (EER) that would result from travelling at that heading.