Post Fly-Away apprehension

[Skeets]

As you may know, last week I had a loss of GPS/spike in HDOP causing my H to temporarily fly away. The H was at 100 feet, and as you can see from the video, there were no mountains or high buildings around that might have caused the loss of GPS. From my view point, the H dropped out of view and I do not think I would have had enough time to touch the on screen "buttons" to take it out of GPS mode and attempt to fly manually. Since then, I have been very reluctant to do any type of flying other than in manual and it being near me. I would hate to think what could have happened if GPS was not regained. The stigma against drones is bad enough.

The onboard video of the fly away is here:

The fly away begins at about 23 seconds

I'm just wondering for those who have had a flyaway or other type non-pilot induced incident, were you a little skittish about flying again, or did you chalk it up to a one in a million anomaly and had no issue with going right back up?

 

[PatR]

Oh man, you bet I was. It took me weeks to put the one I had fly away back in the air after getting it back from Yuneec. The first flight was one of extreme apprehension and kept real close in a wide open area. Two more flights stretching things out further each time relieved a lot of stress. Taking it to a place far away to see what happened next worked out perfectly. The repaired H works perfectly after getting it back. Hope yours goes as well as mine did.

FYI, I haven't used Smart Mode for anything since that fly away back in December...

 

[Steve Carr]

Yup, I'm still timid after an incident on June 22nd. It wasn't GPS related but rather a fault in the ST16. Regardless, my butt cheeks are really sore after I fly. At least your incident had a good ending.

 

[Skeets]

The thing is it was working flawlessly before this (except for the tip-overs, which we all know it WAS NOT MY FAULT ) So there is no telling if or when it is going to happen again. So saying yours has worked perfectly is almost of no point. I've crashed plenty of times with my Helis and planes, but every one of them was my fault in one way or another. This fly away with the H was almost positively not my fault.
I do not fly in Smart Mode except when I am playing around with Watch Me/Follow Me. I think it is much more difficult to fly that way.

 

[Steve Carr]

It was clearly a waypoint error in my opinion. This was where you were doing a waypoint flight, right? It appeared to me that one of the waypoints somehow was in error which sent it to the wrong point.

 

[Skeets]

It was clearly a waypoint error in my opinion. This was where you were doing a waypoint flight, right? It appeared to me that one of the waypoints somehow was in error which sent it to the wrong point.

That is what I thought too. However, I copied and pasted each waypoint coordinates manually onto Google Earth, thinking that one of them would end up some location away from my route. They were all where I had put them when I built the route on the editor in the first place and none were outside of the orbit that I had created.

 

[AH-1G]

That is what I thought too. However, I copied and pasted each waypoint coordinates manually onto Google Earth, thinking that one of them would end up some location away from my route. They were all where I had put them when I built the route on the editor in the first place and none were outside of the orbit that I had created.

I also had way point not cooperating, so I don't use it, to risky and I can see why.

 

[gwhuntoon]

Glad to hear I'm not the only one with a lack of comfort flying the H. A fly away is always foremost in my thoughts which is why I've stayed in Angle mode. I'm always clenching when flying this thing, I hate that. While I know Angle mode is no guarantee I just don't trust the tech. I always keep mine close so I may have time to turn off GPS in an emergency. While I've had no issues, just reading what others have posted makes me extremely uncomfortable flying this thing. Only recently did I try POI mode which worked flawlessly...way cool BTW. But even so, there's this nagging lack of trust I have with this bird.

My heli's I trust, they're always locked in but then I'm in control.

Maybe if by default the drone firmware would just hover in place when there's an irreconcilable error. I know it may drift a bit, but at least it's not flying away like a bat out of h--l. Or, better yet, if there's an error automatically shut off the compass and GPS, then hover in place, and activate an error message via telemetry to the ST16.

 

[Steve Carr]

Even airplanes with redundant systems fail, so why would you expect more from a $1K machine. For the most part these things do darn well. I figure if it works for at least a year then it's done it's job well. Every flight after that is a gift.

 

[Skeets]

But even so, there's this nagging lack of trust I have with this bird.

My heli's I trust, they're always locked in but then I'm in control.

My thoughts exactly. It takes much more skill to fly a helicopter and for me, flying a helicopter is not relaxing at all because you have to have full concentration on all controls the whole time that you are flying, but there is more of a real feel flying a helicopter. When I fly the H, I feel that it is too "electronic" and "artificial". Even with my other drones, even though they are much easier to fly than my helicopters, there is still a real feel to flying them that the H just does not have. I do not NEED to fly the H with GPS. I bought the H because I WANT to fly GPS and it does come with a great camera. I shouldn't have to constantly worry about the GPS failing.

 

[gwhuntoon]

Even airplanes with redundant systems fail, so why would you expect more from a $1K machine. For the most part these things do darn well. I figure if it works for at least a year then it's done it's job well. Every flight after that is a gift.

I totally agree that they do darn well, it's when they don't that's concerning, especially because I have invested nearly 2k, I don't want to loose it or potentially harm others or property. It still seems odd to me given the level of sophistication these drones have, they can't handle what triggers a flyaway better. I still hold to the idea that when GPS is lost or the compass throws errors the system could automatically disable these systems (GPS, Compass, Other), hover in place and attempt to restart them until good data is received while sending a message to the controller -
"Error has occurred, manual pilot control is required!".

To me, not handling this better is what gives the novice a squirrel in headlights look when their drone flies off into the sunset.

In addition, better handling of this has an advantage for the manufacturer in not having to review flight logs and possibly replace their customer's entire drone.

I've heard the math involved in getting a drone to loiter is staggering. I can only imagine what's involved in automating the other functions, so why not better handling of a flyaway situation?

 

[Steve Carr]

I totally agree that they do darn well, it's when they don't that's concerning, especially because I have invested nearly 2k, I don't want to loose it or potentially harm others or property. It still seems odd to me given the level of sophistication these drones have, they can't handle what triggers a flyaway better. I still hold to the idea that when GPS is lost or the compass throws errors the system could automatically disable these systems (GPS, Compass, Other), hover in place and attempt to restart them until good data is received while sending a message to the controller -
"Error has occurred, manual pilot control is required!".

To me, not handling this better is what gives the novice a squirrel in headlights look when their drone flies off into the sunset.

In addition, better handling of this has an advantage for the manufacturer in not having to review flight logs and possibly replace their customer's entire drone.

I've heard the math involved in getting a drone to loiter is staggering. I can only imagine what's involved in automating the other functions, so why not better handling of a flyaway situation?

Perhaps it's a matter of degrees. Total failure is easy to identify. It's the almost fails that are tough to program. That tiny little brain has to sort out the "almost" from "normal". If it sees a zero then it knows that's wrong. If it sees a 39.00052 it has no idea that it should be 39.90052. So the FC says GO to achieve the number. Sooo.....a little off and it's a flyaway. If it's WAY off then it's an error. I suspect the next generation of FC's and firmware will have a new approach to solving the issue. This technology is still in the incubator and just emerging.

 

[PatR]

Good question and one the other guys haven't figured out either. I can say that with military grade birds, that also lose GPS, operators are trained to take over with manual guidance when they lose GPS. There's also a special protocol that automatically takes over when Comms are lost for X period of time. If both comms and GPS are lost for X period of time a bird will enter a programmed maneuver and self destruct. Since the aircraft would not know where it is, could not follow a programmed flight plan, and are not receiving signals from mother and can't be controlled they can't be permitted to fly away. Having long legs they can, and have, end up places they can't legally be. An example is the Shadow found in a tree in Colorado. It was launched in Arizona... Everything failed except altitude and heading hold and only came down because it ran out of gas. They cost more than our stuff by an order of magnitude, and have a lot more room for sophisticated electronics. Just the auto pilot costs in the tens of thousands. Full scale still has pilots in the front seats for this reason. Every airliner must be capable of being manually controlled in the event automation fails, which it does from time to time. From my perspective, the hobby market for multirotors would not exist because of the cost if they were being sold with systems that could do a lot better than they do, and they would be too large to easily transport. There's also the fact there is no certification standard for them, and if there was, once again they would be unaffordable because of the cost to meet higher standards. OTOH, they are a lot better than they were only 4 years ago.

Something I believe is of greater importance is people that fly bigger stuff are in constant training, where that training is mostly about dealing with emergencies. It's pretty well established anyone can fly aircraft that will fly themselves. People just need to know what button to push or switch to flip to make the aircraft do something. The wheat gets separated from the chaff when stuff happens where the automation, buttons, and switches suddenly don't work. People that have trained for those moments instinctively assume manual control and bring the bird home. Those that haven't end up buying new aircraft. I'll suggest practicing flying without GPS as often as possible. What do we do when the gear fails to drop? What if pitch, roll or yaw fails? Could we get it back with the remaining controls? I know I haven't done enough of that and have become much too complacent in my flying.

 

[FlushVision]

...<snip>...
I'm just wondering for those who have had a flyaway or other type non-pilot induced incident, were you a little skittish about flying again, or did you chalk it up to a one in a million anomaly and had no issue with going right back up?

I had a fly away with a Phantom 2 Vision Plus back in April 2015. It was only it's 4th flight. I was hovering about 30' up and about 50' away and, without any warning, it just took off at full speed. I was on high ground on Crompton Moor with only one line of trees between me and the City of Manchester...the direction that the P2V+ was heading in. Fortunately it hit the trees and crashed to earth.

After I had it repaired I asked the repairers to flight test it in their field in Euxton, Chorley, where they gave it the 'all clear'. Even so, I was very nervous of it for several months after. I've not had a problem with it since, but that fly away is still in the back of my mind every time I fly it to this day.

 

[AJupstateNY]

As you may know, last week I had a loss of GPS/spike in HDOP causing my H to temporarily fly away. The H was at 100 feet, and as you can see from the video, there were no mountains or high buildings around that might have caused the loss of GPS. From my view point, the H dropped out of view and I do not think I would have had enough time to touch the on screen "buttons" to take it out of GPS mode and attempt to fly manually. Since then, I have been very reluctant to do any type of flying other than in manual and it being near me. I would hate to think what could have happened if GPS was not regained. The stigma against drones is bad enough.

The onboard video of the fly away is here:

The fly away begins at about 23 seconds

I'm just wondering for those who have had a flyaway or other type non-pilot induced incident, were you a little skittish about flying again, or did you chalk it up to a one in a million anomaly and had no issue with going right back up?

Msg, "this video unavailable"?

 

[Tom C]

I totally agree that they do darn well, it's when they don't that's concerning, especially because I have invested nearly 2k, I don't want to loose it or potentially harm others or property. It still seems odd to me given the level of sophistication these drones have, they can't handle what triggers a flyaway better. I still hold to the idea that when GPS is lost or the compass throws errors the system could automatically disable these systems (GPS, Compass, Other), hover in place and attempt to restart them until good data is received while sending a message to the controller -
"Error has occurred, manual pilot control is required!".

To me, not handling this better is what gives the novice a squirrel in headlights look when their drone flies off into the sunset.

In addition, better handling of this has an advantage for the manufacturer in not having to review flight logs and possibly replace their customer's entire drone.

I've heard the math involved in getting a drone to loiter is staggering. I can only imagine what's involved in automating the other functions, so why not better handling of a flyaway situation?

For the H to simply hover in place when the gps fails would be impossible because it depends on the gps to know whether it moves in lateral or vertical position. Although if Yuneec spent aliitle more time working on a firmware update I don't see why they wouldn't be able to add into the program coding the priority of each technology like for example, if the put in one the obvious of keeping the motors running & in sync, then keeping in contact with all current gps satellites while at the same time always scanning for others with better reception to replace lower reception satellites! Those could be in the Priority #1 category with few other operations then keep making prioritized categories so that when it may have a conflict instead of it thinking that everything is #1 priority which inevitably would make it freeze or in basic terms have an electronic seizure or mental breakdown sending into random acts such as what your did along with random flyaways! Instead it would just refer to the prioritized technologies & see that its 1st priority is to stay in flight & focus on keeping good gps signal & only then after the top priorities are met it can worry about sorting out any other issues which by that point hopefully they've sorted themselves out so you may not even notice it had any issue to begin with!

By comparing dual redundant multi-million dollar planes to as you the others that have replied say a $1k machine, is ridiculous & just sounds dumb. Airlines are engineered to fly thousands of miles each day & have a life span of about 40-50 years (& last I checked its still safer to fly than to drive) the ratio of aircraft's that have crashed because of system failures compared to how many total aircraft that are in service is extreme minor. Intel invested $60 million into Yuneec for the developement of an entire overhaul on their navigation, obstacle avoidance & other technologies to make it so their future drones (which happens to include the H) would be Smart Drones & have the computing power to handle all of the various technologies & programing algorithms that make a drone a Drone.

That is the reason I say that these issues are within the programming code that Yuneec programming engineers write into them. If you compare right now in time to when they first released the H I'll bet there are far less flyaways u other issues as there was at first. Because as the user are put through witnessing their investment flyaway or whatever other issue only when it becomes a large enough issue do the Software engineers at Yuneec Comb through all of their massive amounts of coding & re-write some of the code that may be causing the issues!

What they Should do as a company altogether is after the coding is written for any particular firmware release is to 1st have each & everyone of the engineers comb back through all of it to make sure it does not contain any errors & then only then should they release it out for user to install on their drones! Case in point look at an Apple Iphone Firmware release compared to many android releases, Usually Apple comes out clean running smooth at 1st release & have very few bug issues yet Android would rather get their firmware out faster even if they haven't completely worked out all the bugs. Something to think about is trying to find a way to look at the current firmware coding being as it is android it should not be to difficulty for the right person.

 

[Tom C]

After thinking about how all aircraft have been made to be manually controlled in any case of failures, raises a question about the new AirBus a380's & a few other newer AirBus models because although the divulge that they have dual redundancy systems they are still only using "Fly By Wire" technology ie Joysticks that are only connected electronically to onboard computers, much different then even the Boeing 747's cuz atleast in those they have redundant down to the bone manual flap control etc if it comes done to it. Anyone know what an AirBus pilot would do if he lost power while flying an a380 other than crap his drawers?

 

[PatR]

If we dispense with the term "hover in place" and change it to "hover and drift with the wind" during a GPS loss the conversation becomes a lot more accurate. There's a lot of simple people out there that lack the ability to reason, and will believe that hover in place refers to remain in a stationary hover during a GPS loss. They don't, they can't, and that fallacy needs to be eliminated.

Airbus aircraft can still be flown after the loss of an automated system A larger problem is those in command of them have let their training lapse and often no longer have the ability to fly and aircraft manually. Without a computer telling them what to do they make bad decisions and kill a lot of people. That's how the Airbus flight was lost over the Atlantic. The pilots failed to use the information still available to input the proper commands. Instead they used control inputs that assured they would all die. A Chinese 747 crew did essentially the same thing when they crash landed a 747 at SFO a few years ago. They did not know how to fly manually and failed to perform system checks prior to disengaging the auto land feature from the autopilot.

A common theme with the above is lack of training. Flying requires initial and recurrent training to remain proficient. Proficiency is a requirement of safe flight operations. People think they can grab one of these things off a shelf and instantly be a super "pilot", and create the next cinematic wonder of the world with no previous experience. The real problem is the mindset of those trying to fly these things. They don't understand them, are often not the least bit qualified to fly them, and are 100% dependent on technology to be successful.

 

[fastang]

I would certainly be on edge after any kind of major event such as a fly away. I have the dreaded toilet bowl going on with my H that Im working on fixing with a new GPS unit but it gets your attention real fast when it happens. I came to the drone world from RC helicopters. I've been flying 450 and 600 size helis for 5yrs now and there is no computer to help you hover and you are in control of every movement for every second of the flight. Its kinda second nature to not freak out when something unexpected happens and try to correct the situation. This isnt aimed at the OP but I really believe that most of these problems are people that have 0 experience flying anything manually buy these drones and just expect the computers to do all the work for them. **** it tells you on the box that itll do that right?

 

[PatR]

You heli guys get to work pretty hard at it for sure. Aren't you glad they came up with gyros? I remember the early days of the Hirobo stuff long before gyros came about. The guys learning to fly them could spend 6 months and 15 gallons of fuel just learning to hover.