Aces High Bulletin Board
General Forums => The O' Club => Topic started by: indy007 on May 23, 2005, 12:37:53 PM
-
I know Eagle Tree makes some pretty spiffy telemetry packages, however, what I'm hunting for is an off-the-shelf solution similar to what the US Army is developing for their manpack UAVs.
Is there basically some autopilot windows or linux based software available that works for r/c aircraft? The only thing I really need it to do is orbit an assigned area at max effeciency settings until told otherwise. Been hunting all morning but come up dry so far :(
-
I've built R/C autopilots, and it's awful hard to do cheap. If I were to do it today, I'd use the FMA CoPilot, a wing leveler that uses IR to keep the horizon steady on both roll and pitch axis. I'd have that running and use a microcontroller use a GPS to control the plane with nothing but the rudder. Keeps the math simple, keeps your programming simple. The only think you need to worry about it altitude. You can grab that from the GPS if you want and have it give some occasional elevator inputs. The FMA CoPilot should be able to smooth out the results of any short elevator pitch adjustments the microcontroller makes.
http://plawner.net/4/copilot/fma_copilot.htm
-
Nothin 'off the shelf' with that kinda smart gps position hold and radio downlink available direct to the public that I've seen. There used to be a lil Israeli outfit that had a nice position hold r/c autopilot.. (no downlink) but I ain't seen it for the last coupla years tho.
Homeland security issue?
Assembling some components and a lil original engineering can getcha there. You'll spend about 2 grand to do whatcha wanna do.. and possibly get the attention of some folks you probably wouldn't wanna talk to.
Next alternative is to talk to remote aerial photography outfits in your area.. LOS downlink video setups are fairly common, doubt you'll find autonomous position hold tho.
Aeroviornment (http://www.aerovironment.com/area-aircraft/unmanned.html) did the manpack Pointer system.. might be a good touchpoint.
Hope this helps!
-
Thx guys. Those were pretty helpful. I can downlink GPS info, no problem. Hell, it could be done with those new FRS-GPS radios that transmit their position whenever you hit the key. A spare servo, timer, & 9v battery can transmit its whereabouts to me. I can steal the telemtry from Eagle Tree's laptop software with a memory scanner. I think my main roadblock is transmitting the control movements from a laptop back to the UAV. That's uncharted terrority for me w/o using a USB cable. I figured sticking a mini-itx board on it, with a phidgets usb servo controller would work & an 802.11 link... but if windows crashes so does the UAV...
If you have the luxury of tethering your servos to a USB cable close to you... or are using a mini-itx board & wireless, check out phidgets (http://www.phidgets.com) . I used their controller to build a joystick-controlled remote gun platform. One day I'll eventually hook it up to one of those wireless motion tracking cameras.... or glue a gyro-mouse to my hat and scale the mouse inputs to make it apache style gun mount :) (track IR is overrated!)
Doing it cheap isn't a concern either. It's helpful, but I've got minimal bills, no kids, & just got a 80% raise... so I'm building new toys :) (there's a Specter IR thermal scope with RCA output & a 1" eyeglass mount LCD with my name on it...)
The end goal is to sling these things in the air, and have them orbit singular points and give me a video feed. The new 26 hour paintball format we're using demands real-time intel. I want to stay ahead of the curb and avoid "balloon" reports. And your right Hang... Homeland Security will probably pay me a visit. The cops sure did when I was test firing a 10' long nerf howitzer...
-
Aeroviornment would have what you want.. position hold and direct laptop/control system interface, ala the Pointer System. Give 'em a call! Tell 'em it's for a feasability project.. ;)
Chasin after and operating UAV's takes a ground team.. sucks up time and resources better dedicated to paintball ground ops.
Suggestion.. do a covert ops penetration and install 900mhz CCD pinhole cameras in the trees and along likely routes of approach.
:D
-
Note: as soon as the adversary team discovers yer UAV's he'll get a broad spectrum jammer.. tuning it to 72mhz would be a piece of cake.. and you'll lose yer edge and a lotta bucks instantly. Better to mislead him with a batch of cheap phoney 'intel' balloons and 'leak' yer dependancy on 'em via artful bragging, while yer real video feeds are tree and terrain based. You can even set up very effective ambushes on the routes to yer phoney intel balloons... make him bleed his resources trying to kill a non-existant threat.
muahahahhahahhaaa!
-
Originally posted by Hangtime
Aeroviornment would have what you want.. position hold and direct laptop/control system interface, ala the Pointer System. Give 'em a call! Tell 'em it's for a feasability project.. ;)
Chasin after and operating UAV's takes a ground team.. sucks up time and resources better dedicated to paintball ground ops.
Suggestion.. do a covert ops penetration and install 900mhz CCD pinhole cameras in the trees and along likely routes of approach.
:D
Definately gonna give them call when I get the chance.... I've got a ground crew ready :) All I have to do is maintain a supply of BBQ & they're down for whatever. They're pretty crappy paintball players as it is so no big loss.
We've done the pinhole camera thing. We've done wired camera systems, motion sensor networks, and my personal favorites of motion sensors tied to flood lights (the bastards crawled up and unscrewed the bulbs) they're really only effective against newbies.. who aren't a threat anyways. Nah, the future is on-demand surveillance and some of the semi-autonomous paintball weapon systems I'm tinkering with. If I can remove the human factor, they won't do anything stupid. I'm waiting on public sector technology to catch up really. My ideal scenario is to catch large troop bodies in the open, roll up in an armored vehicle (aka a built up dune buggy), and unleash a wave of mostly autonomous rc cars with cameras & paintball guns. Right now I can't find a cost effective sonar ranging solution for obstacle avoidance. Salvaged units from older Polaroids work great for house robots... but not so much in the field at high speed...
-
Check out the CMUcam. It makes image interpolation a lot easier, plus it can directly run servos to center on movement. It's what I'm gonna use for my front lawn pop-up sentry gun.
http://www-2.cs.cmu.edu/~cmucam/
-
Originally posted by Hangtime
Note: as soon as the adversary team discovers yer UAV's he'll get a broad spectrum jammer.. tuning it to 72mhz would be a piece of cake.. and you'll lose yer edge and a lotta bucks instantly. Better to mislead him with a batch of cheap phoney 'intel' balloons and 'leak' yer dependancy on 'em via artful bragging, while yer real video feeds are tree and terrain based. You can even set up very effective ambushes on the routes to yer phoney intel balloons... make him bleed his resources trying to kill a non-existant threat.
muahahahhahahhaaa!
Actually I think with a clever arrangement of omni-directional antennas or maybe even pringles cantennas... I can probably drive them all with mini-itx motherboards onboard... very low power drain... USB pidgets servo controller, USB wifi, USB GPS, Eagle Tree seagull telemetry... that'd give me everything I think I'd need. I've got most of the stuff to do start writing software... gonna order a Seagull next week... lay it all out, and then find somebody to make an airframe. I can probably get it flying for under $500... it just might not fly for long.
Plus jammers are foridden. They were banned after I jammed cellphones... saw people using them for "secure" comms. Now I pack a bluetooth sniper.
...I used to play normal paintball... then I figured out how much I enjoy EW...
-
Originally posted by Chairboy
Check out the CMUcam. It makes image interpolation a lot easier, plus it can directly run servos to center on movement. It's what I'm gonna use for my front lawn pop-up sentry gun.
http://www-2.cs.cmu.edu/~cmucam/
There's software out there on sourceforge iirc for using 2 CMUcam2's to get stereoscopic imaging. I tried to make a phalanx type paintball gun to shoot down incoming nerf footballs @ > 200fps, but it didn't work out.
-
http://www.uavflight.com/
-
Originally posted by g00b
http://www.uavflight.com/
I shot them an email, hope they respond. That is *exactly* what I'm looking for. Thx goob!
-
(http://www.sonofthesouth.net/leefoundation/civil-war/1861/october/reconnaissance-balloon.jpg)
eskimo
-
Well here's the email I got from uavflight.com
Allen,
Thank you for your interest in UAV Flight Systems products.
We currently are not in a position to support the RC hobbyist, or
non-traditional UAV user. I might suggest you do a search on the internet for other vendors. I would also recommend that you visit http://www.u-nav.com, and http://www.auav.net .
I cannot attest to the suitability of these products, but I do believe they support the hobbyist, and academic markets. The one roadblock you will face is that the FAA has not figured out how to rule on UAVs in commercial airspace, so you are likely to face an uphill challenge to fly legally.
Best regards,
Lee Reep
UAV Flight Systems, Inc.
sales@uavflight.com
http://www.uavflight.com
I kinda resent being called a "non-traditional UAV user" ... seriously... wtf is traditional UAV use? I'd say my use is pretty normal... Predators even carry hellfires now :mad:
U Nav's system is $5000. Can't justify that for only 1 aircraft.
AUAV's system is $1900. That'll probably have to be the option.. just going to end up with a much smaller fleet... Back to seeing if I can't make my $500 mini-itx idea work (or end up with an mp3 playing wrecked airframe).
-
In case you didn’t get my suggestion above, here’s its navigation system:
(http://www.nipa.net/twine.jpg)
That’s less than 1% of what your looking at.
eskimo
-
Using a Mini-ITX, with the weight and (significant) power issues is really unnecessary. Use a microcontroller like an OOPIC, BX-24, basic stamp, or whatnot. The microcontroller can directly manage your servos, so there's no extra weight for servo controllers needed.
If you use the horizon stabilized autopilot from FMA and just steer it using rudder, you don't need more processing power then that. It significantly drops your costs, increases your payload, the time on target, and more.
$500 is ambitiously low, and you've made some design choices that pretty much automatically push you over that hump without even getting started.
If you want to control them and get telemetry, you have a two way radio serial link. Your plane beams back its position, direction of flight, etc constantly. If you want to re-task it, you either give it new coordinates to orbit over the link, and it then uses the GPS to set itself up, or you give it steering directions manually over the link.
You handle takeoff and landing manually, and that saves you a bundle.
-
Originally posted by Chairboy
Using a Mini-ITX, with the weight and (significant) power issues is really unnecessary. Use a microcontroller like an OOPIC, BX-24, basic stamp, or whatnot. The microcontroller can directly manage your servos, so there's no extra weight for servo controllers needed.
If you use the horizon stabilized autopilot from FMA and just steer it using rudder, you don't need more processing power then that. It significantly drops your costs, increases your payload, the time on target, and more.
$500 is ambitiously low, and you've made some design choices that pretty much automatically push you over that hump without even getting started.
If you want to control them and get telemetry, you have a two way radio serial link. Your plane beams back its position, direction of flight, etc constantly. If you want to re-task it, you either give it new coordinates to orbit over the link, and it then uses the GPS to set itself up, or you give it steering directions manually over the link.
You handle takeoff and landing manually, and that saves you a bundle.
Actually I was saying it was cheaper since I've got a mini-itx board laying around & a phidgets controller, plus a borrowed GPS. You were right though, too heavy overall. Haven't worked with an OOPic in awhile. This should be interesting.
-
Whelp.. the other thing you could do is teach your ground team to actually fly the model. Auto stabilization (gyro type autopilots) is cheap. One guy flies, the other mans the camera downlink and camera servo motor controls..
Anything much more sophisticated than that.. and yah hit the two grand number for each airframe.
Dunno how much that'll cost in steaks and beer tho. ;)