i've put some thought into this.
i intend to build an ocean going bot one day but don't plan to start any time soon.
- Energy autonomy (hopefully the ability to remain underwater/at sea indefinitely)
navigation buoys use 2 main methods of power generation: solar power and an Oscillating Water Column.
the problem with both these methods on an AUV is that it is under water.
water blocks so much of the suns power that solar power is far less effective so the closer you are to the surface the better able you are to recharge.
the other method, an Oscillating Water Column, needs to be sitting on the surface where the waves are to bob about to generate power.
surfacing to generate power is fine in calm weather but when the whether is bad your ROV will get blown off course.
- Macroscopic navigation (I heard this was a problem for AUVs, haven't done the research though)
out in the open Ocean navigation is not really too important. there is no land to hit so as long as you are moving in the right general direction surfacing to get a GPS fix once every few days would be fine.
if you are close to land however navigation is far more important as you need to worry about running aground.
off a fairly blank coastline the tide will typically reach 1.5 knots, parallel to the coast at full flood.
if you are close to any sort of headland this will increase dramatically. (the spot where i learned navigation i had to contend with up to 8 knots of tide in a boat that wound average 4 knots.)
on the plus side, any effect of tide will reverse every 6 hours and run the opposite direction for approximately the same length of time so as long as you have not hit anything you will be back where you started 12 hours later.
without going into too much detail, coastal navigation of a AUV with limited power reserves (ie. low speed) would be very tricky.
ocean navigation would actually be far easier.
of course, if you are just going to chuck this in a swimming pool or non tidal body of water non of this is an issue.
- Cost (duh)
so if you are going to let this loose in the sea you have to be prepared to loose it.
the sea is a hostile environment. there is no way you could predict everything that can go wrong.
- A retractable pod that floats to the surface, has a GPS receiver and solar panels to recharge batteries during the day. AUV releases this pod every so often to calculate position deltas and direction of travel (while underwater and "blind", itll be trying to estimate where its going)
moving components increase the likelihood of failure.
far better to surface occasionally to take GPS readings.
- Some kind of built in water wheel that will spin and drive a generator as the AUV moves through the water? So as it is expending energy to move it is also gaining some back (no idea if this is even possible/feasible)
um. no. unless you can design a system more than 100% efficient it would slow the UAV down more than the power it would generate.http://en.wikipedia.org/wiki/Perpetual_motion
- Using a cell phone as a GPS/wireless communication would also be pretty neat, although I'm not sure how difficult this would be
so it is very feasible close to land but if you intend to go far off shore you will have no reception.
satellite phones are an option but very expensive.
try searching for the phrase "Slocum glider" for an insight into how marine researchers have tackled the problem.