[bigiain]

Normal helicopters can autorotate[1] to a landing if the motor stops. A quad is unflyable with one motor out[2]. An octo can self correct for a single motor failure, and is still controllable with two motors out if you get lucky about which of the motors fail second.

[1] http://en.wikipedia.org/wiki/Autorotation

[2] it's probably theoretically possible that a quad with reversible motors (or reverse pitchable blades) could be built to survive a single motor failure, but so far as I can tell nobody, at least in the hobby size quadcopter world, has done so (thoughI'd be surprised if the KMel Robotics and Pennsylvania University research teams don't know exactly how to do it).

[sjtrny]

> but so far as I can tell nobody, at least in the hobby size quadcopter world, has done so

It's not the motors that are the problem. They can spin freely any direction. It's programming the ESCs to be reversible.

Anyway it's already been done a hundred times over:

- Fixed pitch props but reversible ESCs https://www.youtube.com/watch?v=UoReJULASkw

- Variable pitch prop styles exist such as HobbyKings Reaper https://www.youtube.com/watch?v=mAFxkHAiimY and Curtis Youngblood's StingRay http://curtisyoungblood.com/V2/products/quadcopters/stingray....

Autorotation on quadcopters is possible with the variable pitch configuration.

[bigiain]

>> but so far as I can tell nobody, at least in the hobby size quadcopter world, has done so

> It's not the motors that are the problem. They can spin freely any direction. It's programming the ESCs to be reversible.

True - poor wording on my part. And even the ESC programming is a "solved problem" - all the brushless RC cars run ESCs which understand how to run brushless motors in both directions.

So actually reversing the thrust is, as you say, already happening. What I've not seen (yet) is a controller board with software designed to use that ability to stabilise and safely land a quad with one motor out.

[baddox]

There are quadcopters with collective pitch and mechanical linkages to a central motor. They should be able to autorotate in theory (although there might not actually be enough rotor momentum), and not to mention fly upside down. Of course, you don't get the mechanical simplicity of 4-motor quadcopters, and there's still no redundancy for propeller loss.

https://www.youtube.com/watch?v=Scj8_XEEL1A

http://curtisyoungblood.com/V2/products/quadcopters/stingray...

By the way, it's worth noting that there are also quads with 4 electric motors that can reverse so you can fly upside down:

https://vimeo.com/75697349

[TheLoneWolfling]

Counterexample:

https://www.youtube.com/watch?v=bsHryqnvyYA

https://www.youtube.com/watch?v=ek0FrCaogcs

[bigiain]

Ahhh yeah - ETA Zurich. They're the other team besides KMel and PennU that Id expect to have worked (successfully) on this. Thanks. That's _really_ impressive. (I wonder how far away from showing up in the 3DR or DJI gear, or even the NAZE32 or KK stuff?)