Great feedback. Removing ardupilot from consideration.
Wow, that was quick. One comment from a guy who has not even flown it?
I've seen ardupilot go and wasn't all impressed, been more impressed what i've seen from kk compared to that.
So what exactly didn't you like about it? Who did the setup? Do you know that it was setup correctly? Which version of the code? I think it's somewhat disingenuous to make a statement like this based on... what exactly? You saw some guy, fly some code, and...?
Been flying Mikrokopter for two years and there are a lot of good and bads associated with them but one thing they do very well is survive when a motor fails as I've had happen at least three or four times now. Each time I landed with no damage other than a cooked motor or speed controller.
This is, IMO, a very important factor which is not discussed enough.
I tested this with Arducopter on my Octo with very good results as well. It flies with one motor dead... you almost wouldn't know there was a problem. I've actually been trying to figure out how to reliably detect this and signal the operator, because if you can't see the motor is stopped due to distance, you might not know. In fact, I have even flown it with 2 motors stopped. I removed the other prop 135° from the first, to give it a fighting chance. That went pretty well, a bit juddery, but not bad. I then removed a third prop. I actually got it up off the ground like this, but it was hard to control it's path so I landed before I hit something. Still, you could nudge a crash to occur in a particular direction like this.
Yesterday was the first weekend with good weather in a few weeks where I'd have the opportunity to fly the Octo and film it properly with an actual camera operator instead of a tripod in the dark. I wanted to show off the latest of Arducopter and how the Overlapping Octo flies (very well). But first I clamped a Contour to it just to see if I could get usable video. Unfortunately it suffered an in-flight failure, the canopy was not restrained well enough, and during a high speed dive, it flew off. The resulting carnage damaged ~4-5 propellers, and worse, one of the motors rotated on the boom producing a strong yaw thrust. Amazingly, the aircraft actually remained airborne. The problem was this thing is so overpowered, and it was trying so hard to stabilize itself, that the motor thrust was keeping it aloft. I couldn't bring it down. The only thing I could do was actually shut the motors off and let it fall (you'll hear me pulse them a few times), and try and re-engage power to stop it before it impacted. It didn't work as once it started tumbling it was all over.
I hesitate to post the video as it's not exactly great advertising for the system. People must remember I am a developer, not a normal user. I test code, push the limits, and this was also a prototype scratch-built airframe. Many of my videos are of failures rather than successes because I'm usually trying to show interesting failures. As the video says, excuse the jello, the camera was simply clamped to a boom. But if you can see through the jello, you'll get an idea of the stability and dynamic performance of it before the failure occurs.