I'd like to attempt to build a UAV this summer, probably with components from DIYdrones (ArduPilot) since they seem to be the cheapest, but I have a few questions about what I'll be able to get out of it.
For now, I'd like to just plug in some waypoints, manually take off, let the Autopilot take over and fly circuits around the waypoints, the manually land. The aircraft would probably be something fairly slow and docile.
What I'm wondering is how large of an area do I need? Could the Autopilot and GPS keep it within the area of say two football fields, or do I need more room? Also, what sort of altitude range can I get. Will I be able to have it fly around 50-100 ft if I wanted to?
I'm not sure how many people on this forum have experience building and flying UAVs, but I haven't gotten answers to these questions when I asked on another forum.
I'll say now that I am well aware of what kind of regulations surround operation of UAVS. By no means do I want to go spy on people, fly it over densely populated areas beyond line of sight or anything like that. That's exactly why I'm asking these questions. I would like to be able to keep the UAV within my flying field. I'm more interested in learning how the technology works than using the UAV for anything more useful than taking some AP for fun. Hopefully one day I'll be able to apply some of the knoledge to a career in the UAV biz after my finish grad school...
It would be great if some fellow Wattflyers know a little about this stuff so I can ask questions along the way, since "oher forums can seem a little hostile and unwelcoming at times.
That's where I asked first. I did get some great help, but I didn't have as much luck getting answers to my questions above, so I figured I'd try here.
I guess I'll do some more research, but its always nice to get direct feedback. I find some of the RCG threads a little overwhelming, since many of them are a couple hundred posts long. I'll see what I can find. Thanks!
I just read up on your intended product, it sounds pretty cool, I would think that with a good GPS module and a slow aircraft, you might be able to keep it inside the area you are trying for.
Your 3 key parameters are:
--How tight can you program the waypoints, meaning, will the software recognize waypoints so close together
--Flight characteristics of the plane, just because you can program waypoints does not mean your craft in a 3ch config can comply with your directions in the space you want (ie, a rudder turn can go a bit wide and it will take a bit of radius to get to your next point, like driving a semi around a go-cart track) also how fast your craft is flying will make a huge difference.
--Aircraft speed also plays into what I figure is your 3rd parameter. The calculation rate of your autopilot and GPS. Lets say you poll (get your GPS location) every 5 seconds, but you reach your next waypoint before that because of speed or short distance... then what?
Honestly I dont have all the answers you are looking for, but I hope I gave you the right stuff to discover your answer. But I would say, reguardless of how you end up, GO SLOW for the small space so that the autopilot set and plane can comply.
Thanks a lot for the information. Now I have a better idea of how to achieve what I want, and what some of my problems might be. If I go forward with this, I'll try to give updates here and try to explain what I'm doing and how I did it.
Alright, I'm getting closer to buying some components. I've been reading over stuff about Ardupilot and I think I'm getting the hang of it. Some of it is pretty confusing, but it is explained fairly clearly on the DIYdrones site.
I'm interested in Ardupilot because it seems to be fairly simple to set up. Its hard to know how tricky it is until I try it, but from reading about it it doesn't look to bad. I've decided for now not to have a Ground Control Station, which would display data from the flight in real time, such as attitute, altitude, location, etc. This requires a wireless module to transmitting telemetry fomr the plane to the ground, and I just don't feel like buying that right now. I also don't feel like lugging my laptop to the field.
So, as far as I can tell, with the Autopilot I will be able to plug in waypoints using my computer before going to the field. Once at the field, the autopilot records the "Home" position (where I'm starting from) when it's turned on. I can swithch from manual mode (I'm in control) or Autopilot mode using a toggle switch on my TX. Take of and landing is done in manual mode. Once in the air, using a 3-position toggle on my TX, whether to I want to 1. start flying to the waypoints, 2. Fly back to home position, or 3. use Fly-by-Wire, which is manual control with Automatic stabilization. Pretty cool huh?
I made a list of things I'll be needing to get started:
-Ardupilot board: $24.95 -the autpilot
-XY thermopile sensor: $42.95 -for stabilization
-Z thermopile sensor: $24.95 -for stabilization
-GPS module: $59.95 -for navigation
-FTDI cable: $20.00 -for connecting the autopilot of a PC for programming, etc.
-Various pins and connectors: ~$17.00 - For completing the autopilot board
Some of the stuff can probably be found cheaper if I look hard enough, but thats still not that bad for my very own UAV. I'm not sure what sort of plane I'll be using, but I'm getting some plans for a "Mud Duck", which looks like a nice slow, stable flyer with lots of room for payload. http://www.mudduckaviation.com/.
I'm just starting grad school this week and I won't be getting any of my funding or scholarship money for a while, so this might be on hold for a while. I should probably take care of things like rent and groceries first....nahh
Someone on RCG suggested that I go with the wireless telemetry and ground station and work with that before attempting autonomous flight with the autopilot. I think that makes sense, since I can play with it on the ground and make sure the sensors are working correctly (left is left, right is right, up is up, etc). I can fly around in manual mode (or fly by wire mode) and see if it's properly tracking my plane and the stabilization is working right.
The down side is that the Wireless equipment will cost about another $100. But I've been told that it will pay for itself by making sure everything works first and not breaking anything when I jump right into autonomous flight.
Ok, its been a while on this. I've realized that flying an autonomous aircraft is a different ball game than flying RC, and there are some potential issues. My flying field isn't ideal for this, since I am somewhat close to a residential area and a busy road. Also, I currently don't have MAAC (Canadian AMA) insurance in case something goes wrong. Technically, I should get a Special Flight Operations Certificate (SFOC) from Transport Canada before I start flying a "UAV". Maybe its not that big of a deal, but better safe than sorry.
HOWEVER, myself and some buddies at school have decided to build a UAV to enter into some competitions, and we'll jump through all the hoops required to do it properly. There are two major Canadian competitions (UVS Canada and AUVSI Canada) that we're aiming for, and possibly the Outback Challenge in Australia if we can afford it. It's still in the baby stages, so we're not sure what sort of autopilot we're going to use, i.e. open source (Paparazzi/Ardupilot) vs off-the-shelf (Micropilot, $$$$).
I figure I'll report on progess here (or a new thread) anyway in case anyone's interested. However we may get to the point where we don't want to advertise exactly what we're doing since it's for use in competitions after all....
So, anyone want to donate some money for an autopilot, FVP system and an Infrared camera??