We had a great time at the East Bay Maker Faire today, with Jason demoing autonomous ArduCopter functionality to constant crowds. Once the throng thinned out, he started doing stunts like this one, where he drags the ArduCopter away from home and lets it find its way back, terrain following the whole way, and then throws it away like a Frisbee and watches it stablize itself and return home. Later, he proved that it's possible to throw it a little too hard, but we'll save that for another day ;-)
You need to be a member of diydrones to add comments!
Comments
1. Flight log from the demo system
2. Parameter settings from the demo system
Many of us including myself have to second guess if the various signals on a given build are clean enough and as expected without a known good point of reference. Seeing this from a great demo helps us diagnose issues which do pop up from component variability and build proceedure from time to time.
Hope you do not mind providing some reference information as above
It's 100% standard. Just the latest version of the code (2.0.49)
I will be updating this weekend to 2.0.49. The alt hold tracing we have been doing is more electrical than software related but best to keep in step with these great code improvements. I would like to know the actual PID values used in 2.0.49 quad used for the demo. More for a data point check if anything. I love watching the video and seeing the arducopter come back home. Very cool.
@Greg
Sound like you should update to 2.0.49, or at least incorporate the alt hold changes, if you're going to be playing with the Alt Hold.
According the change description:
Added a quick and dirty scaling factor for Camera pitch and roll. (params = CAM_P_G, CAM_R_G)
---> Fixed some bad gain values in Alt hold.
Tweaked gains in Loiter based on feedback.
Made flying from WP faster and independent of GPS update rate.
Flying very nicely today at Maker Faire.
I have the standard Arducopter running 2.0.45. I am using this code at the moment to standardize on for exhaustive sonar testing I have being sharing with several others on the forum (See forum topoc sonar-noise-reduction-update) Hence the reason to known about any specific mods like, power, wiring or different sensor setups on the demo systems. There is a general consensus the sonar in particular is problematic even with the recommended setup in the manual.
Greg: That's just the standard ArduCopter 2.0.49 code (latest version), on standard ArduCopter. We had two going at the same time..
What version code are you using and what frame?
Well I am very impressed. There is not a single flight mode or software combination I have tried so far that could come even close to this. I have to fly my arducopter quad with the radio all the time keeping careful watch, whether in ALT hold or Loiter. I certainly I could not approach the copter grab it and throw it away. Amazing video. Can you please share the code version and exact hardware setup you used? I would be highly interested in the sonar mods / setup or some close up photos.
Gary
You are perfectly correct, we cannot be too careful with all helicopters. Shrouded blades are actually more efficient if the top bell mouth is designed right. I can see the time when all commercial AP stuff will be legislated to have this feature. Open blades in the film industry is a non starter for any close-in shooting.
Great demonstration, and I know you guys trust the code, but motors can burn out, ESC can fail.
Probably goggles should be worn.
Speaking of safety, I guess people would frown on this (I'm wearing goggles.):
Congratulations Jason and acm team.
The Position hold of 2.0.49 seems to be fine. I discharged 2 batery and the ACM not fly more than 5 meter away the PH.
The Altitude hold continue with problem. The baro code seems to be fine. My PID PID need tune but the sonar not work good.
The stabilization is better but continue with small wobble.
It seems need more D.
i flying with small quad (50cm)
-
1
-
2
-
3
of 3 Next