How do UAV's work?

[Tominator]

Can somebody please tell me how all these eletronic thingys and gps stuff work and how much they cost. :D :cool: :confused:

[Rhathid]

There isn't a true commercial UAV pack out there that I know of.

At best you can combine the Seagull system with some other devices.

But, if you wanted to make it work you could start programming microcontrollers and tie into the GPS network.

This is a project I want to work on, as I do program hardware and software. What I want to do is tie into the network, write a program in C or Pocket PC Embedded language (even Java). Using either a Palm or a Pocket PC. The Palm would be cheaper and you can buy hardware to adapt a microcontroller to the Palm.

To get it to work you would have to have some other hardware, for altitude and level flying. That stuff already exists, you should be able to hardware into that as well to take that data, such as altitude, and plug it into the GPS end of things.

So after you have hardware that plugs into the GPS network. (I'm estimating hardware costs for this at about $30-40 if you make it youself using parts from sites like digikey and a few speciality sites). You'll then have to find some good mapping software that will run on a Palm or Pocket PC.

I know that Mapquest.com has a developers license for $1500. With that you get access to everything they use to create their maps. With that I'm sure there is longitude and latitude coordinates.

Logitudes are 69.7 miles apart from each other and latitudes are 49.x (forget exactly). Based on that you can calculate where you are with the help of GPS and your airspeed. Airspeed could be calculated by GPS too.

GPS calculates altitude too, which will come in handy.

Anyway, once you can calculate your position and you know the distance between long/lat using some basic math, you could in theory have a fully functional UAV.

I have been tossing the idea of this around since I first talked to Cam Man about programming PICs. Turns out I have a buddy who is also into it, it's kind of strange, we've known each other a long time and I was on a PIC programming site and the company had endorsed an improvement he made to a popular PIC programmer. (To learn more about PIC's go to www.microchip.com)

If you can accurately detect your alitude and if your plane is level, I don't see any reason why the plane could not also land on it's own if it knew where the runway was using GPS.

I'd be willing to join an online group of people trying to tap into the GPS network and develop UAV hardware/software. :) Any takers?

I don't have much time, but when I do have time I put in 110%. Usually I work on a project from start to finish, so the projects would have to be specfic. Such as, finding out what is required hardware wise to tap into the GPS network.

Best Buy sells a DeLome GPS mapping software for $99. It has the GPS hardware in it, I thought about buying it just to take the GPS hardware apart and figure out how it works so I can make my own.

After that's done, the only other thing is the mapping software. That would take a lot of work as well, but in the end you could do a point and click map of where you want the plane to fly. You'd just plug in way-points.

So we have 2 systems. One could be a ground controlled UAV where the computer on the ground is feeding the plane information and the second would be a plane with everything on board using a Palm PDA or Pocket PC.

Well?? Whacha think?

[Rhathid]

Mission 1 accomplished.

http://www.oreillynet.com/pub/a/wireless/2000/12/08/gps_palm.html

I knew it. I didn't even look into this when I wrote everything above, I just knew it had to be possible.

Okay, based on this, I am not seriously thinking about starting a UAV project. I already own a business, I know how to file the proper business forms and licenses with the government. I'm also somewhat in the know on how to get government grant money for projects like this.

I'm posting another thread on this topic.

[Tominator]

Thanks guys for all that information you gave. A while ago I saw this little to car for like toddlers and babies. It was a sort of a car that I think it was like programable( it was about 4 years ago). It had a few buttons like left, right, backwards and forwards. So you could press left and right and it would go that way. Do you think that would be possible to put that into a airplane?

Note: The commercial was a while ago and it was kind of hard to describe.

[Tominator]

Another thing, Would you consider the pointer made by Aeroviroment a Uav I think its an rc airplane really.http://www.aerovironment.com/news/news-archive/pointsocom.html? Any suggestions?

[Dave Jones]

[QUOTE=Rhathid]There isn't a true commercial UAV pack out there that I know of.

Rhathid
You need to spend a little more time conducting research on the internet.
The AUAV EZI-NAV system is a complete three-axis autopilot system that is commercially available.
The EZI-NAV System is a fully integrated digital autopilot and ground station system that can be used to operate Unmanned Aerial Vehicles (UAVs) with wingspans of 1 meter and up.
It is a complete commercial system all you need to add is the airframe, and if you would like us to we can design and build that for you as well. Check out our web site at http://www.auav.net/
Dave Jones
AUAV.net

[danstrider]

www.micropilot.com
They have a ready-to-fly UAV system.....

Dan

[direwolf]

I know that Mapquest.com has a developers license for $1500. With that you get access to everything they use to create their maps. With that I'm sure there is longitude and latitude coordinates.
The USGS has the Seamless Data Distribution System which provides free access to all public USGS data. Orthographic maps include transform files (TWF, PRJ) that give you most of the data you need to convert image XY to UTM to Lat/Lon. Also, DEM's are provided in the ADF (Arc Data File) format for direct import into any of the Arc viewers (ie. ArcPlot, ArcView, etc). Dozens of other GIS data sets are also available. http://seamless.usgs.gov/

So we have 2 systems. One could be a ground controlled UAV where the computer on the ground is feeding the plane information and the second would be a plane with everything on board using a Palm PDA or Pocket PC.
We're developing basestation software that manages flight paths for multiple aerial vehicles. It also performs all image processing and decision making for fully autonomous missions. Other than the autopilots themselves, the only computers we have on-board our UAV's are Arcom Vipers running Windows CE. We will be using them to relay sensor data and video to the basestation, and to control various systems of the UAVs that are beyond the control of the autopilot computers.

-- Brian Stone

[danstrider]

Brian, Does your company have a website?

Thanks,
Dan

[direwolf]

We're a university: South Dakota School of Mines & Technology (http://www.sdsmt.edu). We're funded by grant to develop COTS (Consumer Off-The-Shelf) UAV systems that are capable of taking on complex recon missions that would be too dangerous or too difficult for a human to do. And we compete in the IARC (International Aerial Robotics Competition). This year will be our second time attending the competition. Here is our team website...
http://uav.sdsmt.edu

-- Brian Stone

[photoguy]

A note about the use of GPS for landing: GPS can get you pretty close, and in line with the landing strip, but it doesn't sound like it would be wise to use it for actually setting the craft down. The problem is both (a) accuracy, and (b) a kind of interference.

(a) On accuracy, the known position envelope is a little like a flattened football on end. The x and y accuracy is better than the z (altitude) value. Don't be fooled by the accuracy statements associated with the marketing effort. These are sampled under perfect conditions, and are more of a statement of the GPS unit's potential. Your indicated altitude will not only be behind by a second or two, but it will often be off by many feet.

(b) On the interference issue: it really isn't interference, but a bouncing of the satellite signal off of nearby objects that is usually the problem. These bad signals can cause some real problems in the short term. For surveying and such, we correct for these bad signals after the fact. It takes minutes to do it. These bad signals are a problem when the GPS antenna is forced to use signals from near the horizon, such as in a turn, and closer to the ground, where signals bouncing off buildings, towers, cars and trees interfere. Look up "pseudorange" for more information.

For really professional applications, high quality GPS recievers, larger antennas, or multiple antenna arrays can be used. Data position smoothing and trajectory prediction can overcome much of the limitations if you have a computer processing the inputs onboard. Real time differential correction is available in many locations.

...but still, the last few feet will be crucial...

Kevin

If anyone really knows a heck of a lot about this, do send a PM.

[sesat]

Not doing trajectory prediction is just downright dumb IMO. It takes a buck of processing power instead of a two thousand dollar GPS receiver with a oven-controlled oscillator.

Antoine lands his Paparazzi craft on the airstrip it took off from autonomously.


Ram.

[Hoylomok]

As for a GPS landing, I plan to make it simple and cheap with a well choreographed parachute landing. Less GPS precision needed with the same outcome. This system would also make emergencies easy to handle.