Discussions

What I want in a nutshell: a UAV airship that will park itself at 70,000 feet or as high as it needs to be to minimize weather impact, keep itself within a 100 meter "box" or so, have enough lift to carry solar panels and other gear, stay aloft for as long as possible, be fairly robust, and be affordable if built in quantity. All done with open source gear to the maximum extent possible.

The reason: this would make for a very nice WiFi relay using high-gain directional antennas, in theory. The same platform would work for cell phones, APRS, and anything else that would normally require an antenna tower or satellite. WiFi needs clear line-of-sight and this solves that problem but good. The purpose is to serve rural areas that lack decent wired Internet access and possibly to provide backup for wired service. I like the idea of moving out to the country but Internet access there tends to suck.

This isn't an original idea of course:

http://www.gizmag.com/lockheed-martin-geostationary-solar-powered-airship/11582/

But so far all I've seen proposed are expensive monsters designed to cover a large region or small airships designed for short-term flights:

http://www.gizmag.com/highest-airship-flight-record/20379/

What about a small airship rigged to cover a county or less? It needs to ignore wind as much as possible so I'm thinking "flying saucer" shape? That would give lots of surface area for a flexible solar array like one of these:

http://www.uni-solar.com/products/commercial-products/pvl/

That company has produced arrays for solar electric airplanes so if the off-the-shelf arrays are too heavy they might have something unlisted that's workable.

Could the UAV be kept stable enough so that directional antennas could maintain reasonably precise alignment? I'd like, say, 3 antennas covering 120 degree sectors within a 10 mile horizontal radius of the airship, probably much less since I like speed and there's the 13 mile altitude to consider. It might make more sense to go with one sector covering a smaller radius and build more drones. Or one antenna to provide coverage, one highly directional one to talk to the uplink. Don't know.

Ideally I'd like to link a fleet of these drones in a mesh network:

http://www.open-mesh.com/

so that if any drones lose their terrestrial uplink, or one simply isn't available, they can still provide service at a reduced capacity. And it'd let the fleet act as an independent intranet. Keeping directional antennas aligned well enough to do this seems like a huge challenge.

I see that the tracking problem has been at least partially cracked:

http://www.rcgroups.com/forums/showthread.php?t=1382551

I like the idea of ground stations being programmed with a list of drone GPS coordinates that provide fallbacks if the closest drone goes offline and automagically searches for the next closest drone. Not sure I could justify the parts expensive if it's avoidable. But the drone's antennas need automatic tracking and stabilization... but maybe not if I skip the mesh network idea and can keep the antennas from rocking too much?

It would be very cool to have a few long-range drone-to-drone links, so that areas far from wired links could get coverage without having to daisy chain a lot of drones and to cut down on intranet latency.

I'm thinking mylar skin with mylar bladders filled with H2. I'd need a ballonet to keep the altitude dialed in? What to use for the frame? It'll need one to support solar panels at the least. Aluminum tubing? Cheap and brute force but maybe too heavy? I think I want a narrow cross-section to decrease wind resistance but could that tip and act as a sail? It'd be nice if it could fly itself into position from hundreds of miles away, otherwise being able to break it down for transport is going to be a big issue.

Probably should have a parachute in the event of catastrophic failure, in the balloon -> load line -> parachute -> payload configuration someone suggested in another thread. High-powered rocket parachute?

It would really suck if the airship found its way into someone's satellite dish path... maybe making the hull out of microwave transparent material would be a good idea?

http://www.balloonkits.com/ and http://www.arhab.org/ look like interesting places to get some materials and ideas.

Heh, too bad you can't put a windmill generator on an airship, and I'm trying to avoid wind anyhow. 

Should I be asking all this over here? http://www.rcgroups.com/blimps-62/

This may all be moot if that distance proves to be too much for reliable WiFi at a decent speed. I'd much rather go with lower altitude and less coverage but I don't see how weather will allow that.

Anyhow, that's enough of a brain dump for now. I suspect that there are very good reasons why no one's already done this?