A Year of Research - Still Not Started UAV Project!

[dracul]

About a year ago I got turned on to the idea of creating a high altitude balloon launched glider which would return to base autonomously taking pictures/video along the way, relaying telemetry to a groundstation similar to:

http://www.members.shaw.ca/sonde/
http://www.gpsboomerang.com/

Little did I know just how much time I would spend researching everything involved and here I am almost a year later with a lot of knowledge but nothing tangible really to show for it.

Every which way I've turned I run into a wall for one reason or another. The main problem is choosing which MCU to go for. If I start with a Basic STAMP I'll have to start creating PCB's, purchase the programmer etc knowing full well that it won't be powerful enough to do everything I need.

I was keen on the Gumstix for a long time, but there again I need to learn Linux and having spent ages looking through their wiki I find it to be incredibly arcane. It's been 14 years since I did my electronics course and I just can't get my head round the info on there. I don't want to have to start hacking things together if I can avoid it either.

I'm currently looking at the robostix as a starting platform. The papparazi project uses it, although I wouldn't use their code because their UAV software has been designed for a totally different purpose. There is a BASIC compiler for the atmel avr platform too which is enticing to me seeing as I really don't fancy having to learn a C language just for this project.

To start with all I want is a unit I can plug a GPS into one end, attach a couple of servos in the other and write some code to read the GPS and move the servos accordingly, but I just can't find a cheap and simple way of doing it.

I feel very frustrated at my lack of progression, I know all the math required for navigation, I'm a skydiver & hang glider pilot so I know how to land somewhere with no power involved, I have an inherently stable airframe, I have working designs for a servo controlled balloon release mechanism, I have the weather balloons, I just need a MCU platform I can get stuck into.

Does anyone have any suggestions for me that I haven't considered? Or perhaps someone has some experience with this already and could lend an ear to help get me off and running?

Many thanks,
Kudos

[workshop]

"I feel very frustrated at my lack of progression, I know all the math required for navigation, I'm a skydiver & hang glider pilot so I know how to land somewhere with no power involved, I have an inherently stable airframe, I have working designs for a servo controlled balloon release mechanism, I have the weather balloons, I just need a MCU platform I can get stuck into."

What about money? You didn't mention your budget. I found it much easier to buy a PICOPILOT than take on the project myself (BasicSTAMP, etc.).

If you skydive and hang glide then I must reason that you're comfortable with expensive hobbies. For about $800-1000 you can go for a PICOPILOT system.

The PICOPILOT NAV2 unit can interface with a BasicStamp (eaisly from the looks of it). It can also read and process the (falling) altitude data from the GSP data stream. The new Propeller chip is a quazi-parallel processor. That's enough power... :)

Think about intergrating exsisting NAV units and just code and solder your unique application...


Jeff

[Tuner]

I am in your same boat 100%.

I am hopping I can use the Paparrazi project as a foundation and modify and build up from their.
My C progamming experience is limited and my linux expereince is null.

I am using Robostix to get me feet wet and soon I may use the CLASIX board from the paparizzi project.

I dont understand why you think the Paparrazi project wont work for you it is generic enough to have no problem doing what you need the only issues I know of is that it will not work with a Helicopter.

I purchased the Atair INU and have yet to get my Controller code installed to test it out but it seems very nice though the price of 2400 is not small change.

I hear you on the Gumstix I thought it was just me their WIKI is awful.

PicoPilot is not a bad idea at all.

I think in the end you would do best paying somone to help you get to a starting point you can manage.

I have just given up on it being simple and realize I will need to learn linux embedded, brush up on C coding...

Good Luck!

[ElectroLawndart]

I am in your same boat 100%.



This boat is getting crowded. :rolleyes:

@dracul
Your project is almost facet for facet the same as what I would like to accomplish. I'm looking at the gumstix/robostix combo. Using GCC on the gumstix for navigation and webcam duties and the robostix for sensor interfacing. I've never used Linux before so I turned one of my old computers to a Linux box just so I can get some experience with it.
I'm also planning on using the Web Special to purchase the Maxstream developers kit for up/downlink comm.

All this head scratching and very little to show for it. Just a couple of sensor projects and a stack of downloaded papers for bedtime reading.

Hopefully when work slows down a bit I'll be able to devote some time to this thing and post some of my results.

It good to hear that the Wiki on the Gumstix does stink. I thought it was just my total lack of Linux experience as the reason it difficult to follow.

It's Saturday night and I have to work tomorrow. At least I have my reading material. :D

-Dart

[dracul]

Thanks for your input guys, it's nice to know that I'm not alone.

I decided to follow all routes to get things started but I'm still running into that brick wall I'm afraid.....

I emailed u-nav.com about the pico pilot several days ago and have received no response from them.......

I decided to get the Basic Stamp Discovery Kit to start working with MCU's and code. I figure that the process of writing the software will be useful even if I end up having to re-code it on a more powerful platform later. I contacted parallax's UK supplier 'Milford Instruments' several days ago and have recieved no response from them......

I'm trying to contact the Paparazzi guys, I've logged in to the IRC channel on dal.net countless times but there's never anyone on there. I've just found an email address for 'Louis' which I'll try now....

ElectroLawndart:

I look forward to see how you get on with the Gumstix/Robostix combo. Please post anything you do, failure or success, it can only help everyone else.

Tuner:

I'm impressed at your monetary commitment! I checked out the Atair INU unit which looks very nice indeed, although 2400USD is more than I'm currently willing to spend on a component albeit a major one. Interesting that on the site they say "Source Code Access Granted Under Agreement", that could be VERY useful to you. I wonder what MCU they use in that unit, do you know???

Workshop:

You mention the new Propellor chip from Parallax, yes it does look very nice and the price is great too. You can run some Pbasic code on it aswell with a library which is a bonus, although having looked at the code examples 'Spin' looks like a much more modern language.

Cheers,
Kudos

[dracul]

Does anyone have any working emails for the Paparazzi guys?