Cleanflight intelligent fail safe?

[RipAir]

hello!

any one ever tried or looked into tuning the fail safe with/around cleanflight?

to get the idea, heres what im thinking of:
since both fail safe variants have theyr disadvantages, it might optimize the whole situation if the procedures would vary depending on exterior parameters
ie. only turn off the props if bellow a certain altitude, and when above do a slow descent to the ground
maybe even with consideration of the current orientation (turn props off when upside down for instance)

thank you!

 

[pressalltheknobs]

It is certainly feasible. The GPS driven "return to home" is an example of a complex "failsafe" automation.

Calling these things "intelligent" or "smart" is misleading since at best they can only be minimally adaptive. That's why "return to home" flies into trees or buildings. That's not an unsolvable problem either but until things like that happen, limitations are often not thought of.

Low altitude is very inexact unless locally calibrated so while you might get something useful it may only work as planned in limited conditions. That's no reason not to try of course. I think it would be interesting. There's lots of research going on so I'm sure someone is working on it somewhere. You might want to check Betaflight... although mostly it's performance driven I think. DIYDrones.com is the other place this sort of thing has probably been explored.

 

[RipAir]

true, since barometers tend to be referenced to sea level, so the pilot would have to set a ground level before take off and/or use radar or similar gadgets to detect the ground (which then again could also be some other kind of obstacle)

but thank you! ill have a dive into those sites

 

[Craftydan]

Rip,

Baros on multirotors are "uncalibrated" devices in the traditional sense -- they are unreferenced at power-on. They are typically referenced off an "arm" value -- 0 is reset at arm . . . and if you've ever had one sitting on your bench while the weather is changing, it can be surprising how quickly the "ground" can move away from a quad without any props on.

While I have a personal disdain for all but the "falls out of the sky" failsafe (anything else is unsafe, but that's my opinion), innovation is the bread-and-butter of the multirotor world. One thing to keep in mind, like RTH, it assumes sensors are available, which having baro is not the norm. Not unusual, but still, most control boards do not have one.

I would also be curious to know how far a quad must drop unpowered before it reaches terminal velocity. I have no data, but my gut feeling is it's a lot shorter than most would believe. Having the props slowly descend and drop from any height greater than that will only allow the quad to drift further away from the point of control loss, since the impact energy would be the same.

 

[RipAir]

ah thanks for the correction! must have mixed stuff up with general aviation

and good thought experiment
id guesstimate roughly two seconds tendencially less than that
so one could only profit from different fail safes, when flying really low
but then again, the damage potential is rather low then too

on the other hand, it starts getting complicated when flying over water
as even a gentle drop could lead to a critical damage
does the drop out of the sky fail safe completely power off?
since that would give it a probability of survival (when dried off properly before powering on again)

what would you recommend to reduce water damage in such situations?
floaters?

 

[cranialrectosis]

Once the copter is down and the motors stopped, my liability is greatly reduced. I want it down and off. I don't care if it's over water. I can build a new copter. I don't want an RTH flyaway. I cannot buy someone a new kid.

I want it down and off.

 

[RipAir]

ofc avoiding tertiary damage is number one priority, incl any kind of life form - except insects, screw them (jk)
im just thinking about number two and three too ... number three being the craft itself and two is pollution