Notices

Definitely a matter of when not if.
I've never programmed for Symbian before, I guess this is the reason why the n900 is the perfect platform for this kind of thing. The full Linux system makes it really easy.
Although a light mobile phone does seem to be pretty well suited to flight instrumentation, if one could get discrete components to communicate with a base station (perhaps the n900 itself!) it would probably be a lot cheaper when it crashes, and probably lighter in the air. In terms of getting the data to the user, a transparent hud seems like a good middle ground between regular flying and fpv flight.

I wonder how tricky it would be to get a hud system like that working; again, all the hardware is here, just easy software

Checkout DIY Drones.

A n900 powered autopilot certainly would be interesting. It would have to be made fairly reliable though.

Slightly offtopic. This could lead to incredible n900fly scores though.

Autopilot would require the N900 to be able to send servo control pulses (which would require USB host mode and some sort of USB->servo thingy [these can be built form microcontrollers easily but unless you have old PIC16Fs laying around you probably would want to go with more advanced uC and then you could go to full arduino and get all you need [except python...] without using the N900).

Make no mistake: I appreciate the geek factor of doing something like this with the N900 but it's not really an optimum (or even sensible) approach to solving the autopilot (or datalogger, or telemetry) -problem. But not everything is about the optimum or even sensible solutions, look at TCP/IP over avian carrier for example...

I'm afraid you're probably right.
If a n900 to servo system could be implemented, it could be a quick testbed for ideas though. Cheaper (providing one already owns the n900) and much easier to program. A fully set up platform.

Dunno, but the requirements for this app seem to be pretty close to a 'rollercoaster tracker' app I've been wishing was available for the N900.... Track speed, altitude, video... render the whole thing in a '3D replay'... upload to the cloud... share the experience...

Develop in Qt and it should run on (some) Symbian and N900.

A relevant thread: http://talk.maemo.org/showthread.php...258#post675258

And something on HUD hardware: http://augmentation.wordpress.com/20...than-expected/

In summary: there is no see through consumer device yet.

Edit: I'm also in email disucssion with Vuzix europe representative trying to coax some info out of him (like will the display driver for the non see through stereo glasses totally freak out if one display is removed [which would allow a "glance" monocular "HUD" with one display])

I managed to fly one N900 using kites, for KAP (kite aerial photography). It was attached to the line of the kite in a rig with two servo motors (for pan and tilt). The servos were controlled by and Arduino with a Bluetooth modem, which received the pan/tilt commands from the N900. It also streamed video to another N900 on the ground, which in turn sent pan/tilt commands (created using just the accelerometers, by tilting the device) to the N900 on the kite. The Wi-Fi signal (in Ad-Hoc mode) was very good for about 150m, but after 160m or so, some video frames were being lost. For KAP purposes, the video was still good enough for about 200m or more, but I would never even try to control a model aircraft with an N900 beyond 100m! A router with more power and a good antenna could get you a wider signal range from the ground to the device on the aircraft, but it might not help as much the other way around...

This project was part of the Nokia Push contest. You can see the official blog for more information (under "More KAPing with the N900 Updates"). There is also a set on Flickr with images (and even some code) from the project.

I like the way you people think but I have only one question: Isn't this a bit overpriced? 2 N900s is like 1200E, I'm pretty sure an ATMEL could be used for controls and radio (add a radio), and a small camera could be a lot less and take (maybe) similar images.

Streaming video could also be done cheaper via analog, I assume.

Is this really the best solution out there in bang per buck?