| > Thanks for the argument, because I am new to model flight, actually
| > I am not a model pilot but a computer& electronics hobbyist working
| > together with some model aircrafters, for me it is very intersting
| > to have things go - automatically.
Ok, and some would be interested in it, but it's not likely to replace
`wiggling the sticks' ever.
| > First thing to get working is a device which would fly the plane
| > in a certain route, say a few miles south and back again.
| > then there is no radio control possible with standard RC transmitters.
| > Why not use a GPRS telephone modem to control the plane
| > as well as transmit telemetry and GPS data.
As long as it's 100% reliable (or at least as close as our gear is)
and doesn't have per minute fees ...
We usually fly our planes visually, so there's no need for more range
than our eyes give us. Really, the only big improvement that we need
in radio gear for our current purposes is spread spectrum, so we can
1) not have frequency conflicts and 2) not get knocked out by a single
strong signal. (Though an autopilot of some sort might be useful if
the signal is still lost somehow -- perhaps cut power, keep the wings
approximately level, keep the nose level and do circles? But is it
worth the weight and cost for most?)
| > There is of course the question of responsibility, it will create
| > legal problems to have a model fly out of sight, although it does not
| > mean that it is out of control. This is definitely new stuff for
| > hobbyists.
It's not new at all. People have done this for quite some time. I've
even seen shows on the Discovery Channel and the like. The most
recent one I recall was the one where teams were making autonomous
helicopters to survey a site for bodies or something in some contest.
| It's not very common, but I wouldn't say it's new. There have
| been a number of GPS auto-pilot projects (included several
| attempted trans-atlantic flights).
... and several successful ones. Maynard did it with an 11 lb plane
(http://tam.plannet21.com /) but his plane certainly wasn't the first
autonomous plane to cross the atlantic, only the first to be small
enough to be a `model' under FAI rules.
There are even off the shelf navigation systems available. I see ads
for them in the back of the modelling magazines from time to time.
| I think your chances of automating a normal landing are
| pretty small.
Depends. If you have a large field to land on and don't need anything
precise, a GPS could do it. Just keep the wings level (or almost
level), and don't let the nose go down or up, cut the power so it's
not enough to maintain altitude, and use the rudder to turn circles.
As long as it's over the field when you do this, and your circles are
tight enough, it'll land on the field.
Beyond that, you could come a lot closer to a `conventional' landing
with just a little smarts. GPSs aren't really accurate to more than
10 feet at best so it won't be a high precision landing, but as long
as you keep the plane level front to back and side to side, it's
simply a matter of controlling the rate of descent/ascent via the
throttle. `Aim' the plane at the field with an acceptable altitude
and distance away, and then have it narrow in on the target by
adjusting the throttle (and making any small adjustments to course
Doug McLaren, firstname.lastname@example.org
If something's hard to do, then it's not worth doing! --Homer Simpson
Click to see the full signature.