kill switch and 'chute to prevent flyaways

[roykirk]

I'm a long time UAV enthusiast, but only little manually flown devices. I've been seriously considering moving up to the Phantom 2 in the next few months. One thing that keeps giving me hesitation is the fear of the much debated "flyaway." That's actually secondary to my concern about whether it's going to take out some kid when it finally decides to crash.

So on this topic of helping "stop" a flyaway in its tracks (if you react fast enough), I had a thought. Forgive me if it's dumb or has already been discussed. Several manufacturers are now selling parachute recovery systems for quadcopters that can be remotely activated through an attached servo. Being somewhat of a hobby electrician most of my life, I thought it would be pretty easy to wire a relay to that parachute servo that would then run to a separate switch that would immediately kill the battery supply to the rotors. How it would work: Quad pilot notices a flyaway may be unfolding, or they've otherwise lost control of their unit. If they can react fast enough, they hit the parachute switch which deploys the parachute while at the same time cutting off the battery. Brick floats safely to ground. Theoretically.

On one hand this kind of feature shouldn't be necessary as you're supposed to be able to brick the Phantom remotely, but some of the flyaway reports I've read have reported this may not work with all situations. Anyway, that was my late night thought. Feel free to tell me it's the stupidest idea in the world, or that 20 other people have already thought about it and are marketing it.

 

[doug86]

Well, the definition of a fly-away is loss of control, generally attributed to a failure of the radio link-up between the radio control transmitter and the receiver on the aircraft. So, assuming that there is no way to use the radio equipment to fly the thing home, there would be no way to tell it to do anything else, like turn it off or deploy a parachute.

I'm not aware of any feature that allows us to "brick the Phantom" remotely. You can choose to manually activate the Return to Home feature, but that doesn't "brick" anything; once the Phantom is back in range, you can generally take back control. The phrase "brick" in the tech world means "render useless".

 

[Hughie]

Really pleased that you are looking at this.

You would presumably need another channel to do this ? Is this possible ?
As I understand it by definition, flyaways put the pilot in a position where they are not in control. If the Phantom flight software is ignoring any RC inputs, this extra channel will have to be operating independantly of the extisting radio link. How would you do that? You might need to run something on a completely different frequency.

Interesting though.

 

[roykirk]

Sorry, I should have been more clear what I was thinking. I'm envisioning a switch/transmitter completely separate from the quad's control transmitter. That's what could be the beauty of it if it would work. Your Phantom may be refusing to accept any input for whatever reason and is trying to take off on its own, but if you have a separately wired switch (controlled by a separate transmitter) that will cut off the battery supply, you're effectively bricking it completely independent of the Phantom's native control system. Yes, you'd have to take some things apart to install such a switch between the battery and the motors, but it certainly can be done. And with the parachute being controlled by the same separate transmitter, you'd be covered there as well. The only thing this idea doesn't cover is if your Phantom is out of sight and/or you don't have FPV and you have no way of knowing its running away. What brought up this idea was in reading flyaway accounts from people who had long enough to realize they were about to lose their Phantom to a flyaway. They had enough time to try several fixes to regain control but none worked. In such situations if you could go to this emergency kill/parachute switch, you might be able to stop a flyaway in its tracks. That was my thought anyway.

Since my original post, I've discovered there is a company that is indeed marketing something similar, but on a quick cursory glance it appears the parachute/kill switch are part of the quad's radio transmitter. That's what I was hoping could be avoided. If things are so messed up your Phantom won't accept any input, is a parachute deploy/kill motor order going to work? I'd say if it's a completely separate system you're going to have better luck.

 

[jadebox]

The likelihood of a fly away is very low. The likelihood of a kill switch and parachute system working when you need it isn't very high. Therefore, the chance of the system ever being used to save your 'copter is very very low.

Adding a kill switch adds a chance for the power to be disconnected when you don't want it to be. Adding a parachute introduces the possibility of the parachute deploying accidentally. This means that there is a greater chance of something going wrong.

So, I think that you'd end up with something that's more likely to cause a problem than it is to prevent one.

-- Roger

 

[IrishSights]

If you have a P2 fully loaded with iosd mini, gimbal, camera and maybe a GS air end you are at the Max laden weight. You then have to add a 2nd reciever, swtich/servo and the parachute itself you are going to be grossly over weight bringing its own inherited problems. On a larger multirotor with much bigger payload its maybe a goer, but not for a Phantom.

As far as I can see its a non starter, but hey...good to talk!

 

[roykirk]

If you have a P2 fully loaded with iosd mini, gimbal, camera and maybe a GS air end you are at the Max laden weight. You then have to add a 2nd reciever, swtich/servo and the parachute itself you are going to be grossly over weight bringing its own inherited problems. On a larger multirotor with much bigger payload its maybe a goer, but not for a Phantom.

As far as I can see its a non starter, but hey...good to talk!

Very good point. I just know I have seen videos of Phantoms outfitted with parachute systems and since the added switch to kill the batteries would weigh, at most, a few grams, I thought it might be doable. But those Phantoms may have been stripped down models too. I didn't look that closely. Probably just being overly paranoid on my end. As I was reading these many experiences of flyaways, I started to feel the same horror and anguish those people did as they chased their Phantoms on foot or by car trying to regain control of them, only to have them eventually disappear from view. The answer is that I likely need to stop reading such stories.

 

[IrishSights]

After 10 months flying a lot I have never had a failure or Flyaway. Postings usually only occur if there is a problem so giving the false impression that problems are widespread.

Fly carefully, within the law, with a checklist before and after every flight. Also as part of preflight always do a risk assessment and don't be shy about deciding not to fly. Understand your craft, safety and the environment including the weather.

 

[oscar19]

Maybe the trick is NOT to have the operator "pushing a botton". Maybe the real solution is to use the app pre-set parameters go into a failsafe mode when certain things happen ie go outside pre-determined range or a change in speed indicating a plummet from the sky.

This solution would preclude addition hardware for a separate radio signal. The trick would be to change the app to operate this chute servo independently of the operator.

 

[The Editor]

How do you propose to 'cut off' the battery supply?

If we are talking a larger craft running say a couple of 10,000mah 6s packs then you will need to control/circuit break something that can supply in the region of 600amps !!!!! I would not want to do that through any sort of relay circuit since the contact gap would be very small on anything lightweight. Move up to more meaty liquid filled relays (to prevent sparks/arcing) and you have a circuit breaker that could weigh more than the aircraft :shock:

 

[roykirk]

Maybe the trick is NOT to have the operator "pushing a botton". Maybe the real solution is to use the app pre-set parameters go into a failsafe mode when certain things happen ie go outside pre-determined range or a change in speed indicating a plummet from the sky.

This solution would preclude addition hardware for a separate radio signal. The trick would be to change the app to operate this chute servo independently of the operator.

The one commercial product I found earlier advertised that it did just that. The operator could manually kill the motor and deploy the parachute and it also had certain auto settings where it would shut itself down and deploy based on certain programmable parameters like sudden loss of altitude, speed, etc. My idea sounds like it's probably not worth pursuing, but it sounds like others are already working on similar ideas.

 

[ianwood]

There are many reported flyaways that aren't actually flyaways (e.g. user loss of orientation, inadvertent inputs via RC, etc.). Most actual flyaways are caused by IMU/compass errors and most of those are brought on by user error.

The P1 and P2V may (or may not) have radio link induced flyaways but the P2 certainly does not.

To safeguard against a real flyaway situation (e.g. stuck gyro), flipping to manual mode is the simplest way to stop it from disappearing.

 

[Happyflyer]

Interesting idea. I like the one of total power loss, quad falls fast, sensor detects fall, and chute opens.
With my bad luck the quad would land without any damage on top of any 80 foot tree. :roll:
It would be slick to see it float down slow instead of splatter.

 

[Meta4]

Parachute systems already exist and are sometimes used on large hex and octo multirotors but their size and weight makes them unsuitable for the Phantom which is already close to its max flying weight. Here are some http://www.fruitychutes.com/buyachute/u ... dles-c-21/

You also have a problem fitting this to a Phantom. There isn't much room between the rotors and that's before you consider blocking some of the GPS signal.
Recent discussion with DJI people confirmed that parachute systems are too big and heavy to use on the Inspire so the Phantom isn't going to be carrying any.
Concentrating on horror stories magnifies the situation out of proportion with the real situation. You can't lumber a Phantom with equipment to deal with every imagined scenario. Much better to learn how it works and what could go wrong and have strategies to prevent these situations developing.

 

[IrishSights]

Parachute systems already exist and are sometimes used on large hex and octo multirotors but their size and weight makes them unsuitable for the Phantom which is already close to its max flying weight. Here are some http://www.fruitychutes.com/buyachute/u ... dles-c-21/

You also have a problem fitting this to a Phantom. There isn't much room between the rotors and that's before you consider blocking some of the GPS signal.
Recent discussion with DJI people confirmed that parachute systems are too big and heavy to use on the Inspire so the Phantom isn't going to be carrying any.
Concentrating on horror stories magnifies the situation out of proportion with the real situation. You can't lumber a Phantom with equipment to deal with every imagined scenario. Much better to learn how it works and what could go wrong and have strategies to prevent these situations developing.

+1

 

[justin00]

http://www.digitaltrends.com/cool-tech/ ... ash-style/

doesn't look like its for our 'cheap' stuff though

 

[BaldEagle]

Having a chute doesn't necessarily guarantee a safe landing, right? The chute won't be controllable, right, so trees, roads with cars, water, and buildings, and other things could lead to trouble. It seems to me that deliberately putting something onto one's craft that does not allow for a controlled descent might also be problematic, right?

 

[roykirk]

This is more in line with what I was thinking: http://www.marsparachutes.com/mars-mini/

At 3.6 ounces, it seemed like something that could fit as far as weight anyway. But as Meta4 noted above, mounting might be the biggest issue.

 

[Happyflyer]

Now that is still a slick little chute. I would not call that cheap by any means.

 

[PhantomFanatic]

The concept is certainly feasible. I would cut the motors and delay perhaps two seconds, as you need sufficient airflow for the chute to deploy properly.

But, you would need another receiver, for that channel and another transmitter. The main problem might be the weight of the relay as its contacts will be cutting a lot of current. But, 12VDC relays, for automotive use, might be small and light enough. You will need to put a small capacitor across the contacts to reduce the high voltage spike, during disconnect.

I, too, have seen the drone chutes advertised. I don't know how they would attach to the top of the phantom, but that is doable too. If your bird is out of range, though, your other transmitter and receiver need to still have contact. Perhaps a frequency with more range, or higher transmitt power, longer antennas, etc. (If legal.). If not, get a ham license!

Your other problem is if it is out of visual range, you need to be able to find it. There is a thread here about GPS trackers, that will lead you to the drone. I have one on order, that I bought from ebay.

Have fun! It sounds like a neat project, from a design perpective.