Sunday, July 26, 2009

FlightGear Interface Test - Inbound

If the data goes out, there must be a way to pump it back in.


Still network interface so far, but all the pieces to start a safe and crash proof development and testing are here. A couple more programing exercises first though ...

Friday, July 24, 2009

Sensors Test - Quick

Quick app to list and test the sensors.

Probably miscoded, but hacked up as it is, it is doing something. With an added bonus of Orientation Sensor, tested on a train, reports only couple degrees variation between full stop and speeding on a straight track. And even that might be a consequence of the train's suspension.

Either way, orientation problem solved before it even started!

Wednesday, July 22, 2009

FlightGear Interface Test - Outbound

It works.

Over the network interface for now, but FlighGear happily provides sensors found in G1 phone.

A small network issue later and it worked for real.

Nothing useful has been done with the raw data for now, so basically nothing to see here really.

Tools of Choice - FlightGear

Probably the safest way to learn about aviation & programing without crashing hazard.

Things to do with FlightGear include:
  • creating an interface with Android device (network and serial)
  • simulating G1's sensor telemetry
  • possible modification or creation of RC plane's flight dynamics for simulation
A few virtual flight lessons aren't excluded before starting on anything even remotely resembling a functional RC/UAV auto pilot.

Tools of Choice - Android & G1

It's always good to learn about up and coming platforms right?

Android is one of the things I wanted to try, so the next logical step was a hunt for HTC's G1 device (for now, kind of an expensive rarity in my neck of the woods).

Nifty little gizmo, so what will be done with it:
  • internal sensors for flight control - GPS (position and speed), digital compass (direction and navigation) & accelerometer (aplied forces and hopefully plane orientation)
  • Java on Android for computing, control (serial) and telemetry broadcast to ground station (GPRS)
  • probable "rooting" and hacking for control over serial console
  • possible directional wi-fi antenna for test flight tracking and debugging
First post might have jinxed the project a little, since I just figured out that so far there is no support for serial communication trough Android API's. So much for choosing the right tools to fit the job, but any other way wouldn't be as exciting.

Though the hacking of the phone will come a bit later, would like to play with it for a while before I mess it up completely.

Tuesday, July 21, 2009

Goals and Conditions

So, this will mostly be a learning experiment, because I need a constructive excuse for new toys! And what would be more constructive then learning something new?

But to be fair and to treat this as a real project, some thoughts have gone into what to expect from it ...

Goals are the following:
  • learn FlighGear simulator
  • try out RC plane modeling / flying
  • learn basics of avionics and navigation
  • learn Android programming / hacking
  • brush up on electronics and signal processing
  • figure out what else is needed to make this project successful
Here it might be good to stress out that this is not necessarily the order, since the execution is more or less free for all. And how are we going to judge it a success?

Conditions that apply:
  • acquiring any new and beneficial skill or knowledge
  • having a blast (it's starting as an vacation project after all)
  • possibly in midst of all this creating something shinier and more interesting engagement of my time
  • controlling device and/or flying contraption spontaneously combust or disappears in a hole of space and time
  • result being a flying contraption of sorts, capable of navigation on it's own with an added bonus of being able to take a few awesome pictures
Any of these and the project will be considered time well spent! As from now on, anything goes, see what happens.

Monday, July 20, 2009

Prolog

I always wanted an RC plane!

And now an opportunity (financial, with sufficient amount of free time and knowledge) presented itself to do something about that.

And with it, this train of though:

Cool thing these days is to put a camera on it, right? I do admire a good aerial photo, but the closest thing to a camera I have is my cell phone (I developed an unhealthy habit of taking all my photos with it, because in this fast paced world, if you know what you're doing, 3 MP's should be enough for most occasions). Now, since I would already have a device on it about as powerful as my first laptop, wouldn't it be even cooler if it could be made to fly the plane itself? And to make it interesting, make it fly with only the sensors in the phone!

How *hard* could it be?