prop loss failsafe recovery

Post Reply
tovrin
Posts: 705
Joined: Tue Sep 20, 2011 4:08 pm

prop loss failsafe recovery

Post by tovrin »

I know i dont watch the IDEAS forum often, and I hope I'm not posting something already talked about, i did 'scan' the topics first...

http://www.suasnews.com/2013/12/26266/e ... ller-loss/

the above link has a video and writeup about a project they are working on to recover from a loss of prop on a quadcopter. on Hex and octo i have heard of recovery due to having enough props to maintain and land. but on a quad this has been nonexistant. ETH uses a method of what i can only call 'tailspinning' to safely land the copter during loss.



I swear i saw this on TED once, is this something that could be implemented, at least maybe on the STM32 platform (where there is memory to spare), even if an aux switch had to be flipped to engage it?

Sebbi
Posts: 478
Joined: Sun Jul 08, 2012 1:08 am
Location: Germany
Contact:

Re: prop loss failsafe recovery

Post by Sebbi »

Wow, that would be really interesting to have ;-)

User avatar
treym
Posts: 258
Joined: Sat Jul 21, 2012 12:28 am

Re: prop loss failsafe recovery

Post by treym »

This technology is patent pending.


but .. you can find "how to make it happens" in this doc : http://students.asl.ethz.ch/upl_pdf/337-report.pdf .

Sebbi
Posts: 478
Joined: Sun Jul 08, 2012 1:08 am
Location: Germany
Contact:

Re: prop loss failsafe recovery

Post by Sebbi »

Can you please point out the page where the author discusses recovery from a 1 rotor loss in a quadrocopter configuration? I couldn't find it right away ...

User avatar
EOSBandi
Posts: 802
Joined: Sun Jun 19, 2011 11:32 am
Location: Budapest, Hungary
Contact:

Re: prop loss failsafe recovery

Post by EOSBandi »

I have serious doubts about it. ETH guys are geniuses but their setup always includes a very expensive motion capture setup with off-board processing.
On the demo you can clearly see the MoCap reference balls on the copter.

frog32
Posts: 55
Joined: Sun Nov 20, 2011 9:39 pm

Re: prop loss failsafe recovery

Post by frog32 »

The failsafe controller uses only hardware that is readily available on a standard quadrocopter, and could thus be implemented as an algorithmic-only upgrade to existing systems.

So it should be doable, but they have arm processors as far as i know on their quads. I once attended a talk about them and they said it's the equivalent of a modern smartphone. This was last year.

User avatar
Hamburger
Posts: 2578
Joined: Tue Mar 01, 2011 2:14 pm
Location: air
Contact:

Re: prop loss failsafe recovery

Post by Hamburger »

Juyt curious
How relevant is prop loss in real life with proven configurations?
What exactly is "prop loss"?
I can think of (from personal observation)
. A short in motor resulting in fire
. Bearing blocking
. prop loosing one blade
. Esc no longer driving motor
. Breakage / disconnect of one of the involved wires

Some of these cases may equal to a simple reduction by one drive train while in most cases I expect nasty side effects like deadly vibrations or instant lipo destruction

User avatar
Crashpilot1000
Posts: 631
Joined: Tue Apr 03, 2012 7:38 pm

Re: prop loss failsafe recovery

Post by Crashpilot1000 »

+1
That is reality.
Prop loss can be avoided completely.
BTW: If someone wants a very crash resistant quadrocopter I would suggest to go for a Walkera Ladybird....

Sebbi
Posts: 478
Joined: Sun Jul 08, 2012 1:08 am
Location: Germany
Contact:

Re: prop loss failsafe recovery

Post by Sebbi »

"How relevant" ... well it never happened to me, but I bet pilots would feel saver if their quads could land relatively safely in the event of a prop loss (or damaged motor). It's not the same as the redundancy provided by hex or octo configurations, but hey ...

User avatar
Hamburger
Posts: 2578
Joined: Tue Mar 01, 2011 2:14 pm
Location: air
Contact:

Re: prop loss failsafe recovery

Post by Hamburger »

I would not want to stop anyone from pursuing this route. In fact it made me think how relevant as in how high percentage were those events where some smart algorithm was able to recover from. From personal experience the probably non.recoverable prop loss errors outweigh the recoverable ones.YMMV

balrog
Posts: 3
Joined: Tue Apr 30, 2013 5:08 pm

Re: prop loss failsafe recovery

Post by balrog »

So while the whole topic is a little sillly, in theory this stuff is already implemented in MultiWii, Baseflight as well as most other flight controllers that use PID loops. All you need to do is make sure your roll and pitch integral factors (the I in the PID) are sufficiently higher than the yaw I, and that there's no clamping of the integral. In theory (because I don't think anyone's actually tried it), this is enough to get the same behavior as in ETH's video.

Now detection of this event and switching to some sort of head-free mode to keep the quad steerable is a different topic (not really difficult either).

User avatar
linuxslate
Posts: 91
Joined: Mon May 13, 2013 3:55 pm

Re: prop loss failsafe recovery

Post by linuxslate »

I have had 2 crashes due to broken props in flight.

These were the only 2 crashes that resulted in breaking arms, or any expense other than just broken props. (Last one bent a motor shaft, too.)

My solution - Buy better props; Pre-flight Quad.

Post Reply