What's our vision for a more autonomous mission planner?

[rolker]

Do we take baby steps with the Survey Pattern element adding features such as this suggestion from @valschmidt ?

"Another feature (not critical) would be to a survey pattern whose line
spacing grows linearly across the survey. One would specify the rate of
line spacing growth as a parameter. Useful for mapping slopes."

Or, do we take a higher level approach where we specify a polygon, and let the vehicle (or team of vehicles) figure out how to best map the area?

[valschmidt]

This is a good question. I think both.

Damian's work can do the polygon approach and at some point I hope to re-implement it on our boat, but this is a relatively low-level on-board capability for which the mission planner need only set a behavior and specify a polygon. Many boats will not have this kind of low level capability and so the mission planner should also be able to create a number of creative, mission specific but somewhat "manual" survey patterns.

Here's a list of some:
Variable line spacing lawnmower.
Alternating line spacing lawnmower (for example for PMBS sonar's that don't cover nadir well)
Expanding box pattern (most efficient search)
Sliding box or racetrack pattern. (Useful when you need tight lines but don't want tight turns.)
Reference survey pattern (2 orthogonal, overlapping lawnmower patterns)
Polygon, start point + behavior inside the polygon (for algorithms like Damian's)
Zigzag line pattern (for example to map an oceanographic front)
Patch test survey pattern(s). (There are many ways to do this, but I like to find a flat spot with a steep edge. Then I run 3 lines positioned over both the flat area and running up onto the steep edge, spaced 2 max water depths apart, each run in both directions, the center line run at a slow speed and a fast speed. )