Please select Into the mobile phone version | Continue to access the computer ver.
Another drone ALMOST lost in Iceland - Sudden lost GPS, compass, etc
12Next >
4734 46 2017-9-27
Uploading and Loding Picture ...(0/1)
o(^-^)o
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

Well, the day had to come where I would write this, as it seems a big number of Spark users need to constantly deal with unpredictable problems while trying to fly their drones in demanding conditions. I had a terrifying flight in Iceland where I almost lost my drone, following a sudden GPS loss, Compass error, NFZ zone and all sorts of errors which made the Spark drift away madly. Fortunately it kept video feed, and I was able to return it in Sports mode, even though it looked light it wasn't properly responding to the RC inputs. As soon as the GPS signal was lost, I lost sight of it in a matter of seconds and all sorts of things crossed through my mind, mostly things like "this is it, it finally happened, another lost Spark". I managed to keep my cool and saved it using Sports mode, rising it's altitude a bit and then trying to return it through an extremely unstable video feed, where the AC didn't feel like it was properly responding to my RC inputs. Words are not enough to explain what if feels like to see a several hundred-dollar device suddenly disappear from sight and apparently not respond to any control, when it was working perfectly the second before...

I know many people speak about user errors and, as a newbie myself, I bet I constantly make a lot of errors while flying. Still, I wonder how all of a sudden appear a considerable number of posts on this forum speaking about lost GPS, compass errors, etc, after the last firmware update... Since I've made the last compulsory update, while doing some tests at Portugal, I started having an IMU calibration request, every time I turned on the Spark. I read that this should not happen that often, so I always decided to turn the Spark off and on, and then the IMU calibration request would disappear. This was a first sign that worried me, but I kept flying.

One of the main reasons I bought the Spark was to fly it in Iceland. I'm a professional landscape photographer who is addicted to Iceland, and having the possibility to film some sceneries from above was something I always wanted to do. I knew that Iceland is a very demanding place for a drone, but had no idea I would almost loose it because of a series of sudden unexpected errors. Fortunately this happened on the last day of the trip. Before that I flew it with low temperatures, cold batteries, very high Kp's, strong winds, fortunately having no problems.

I'm giving you below Airdata flight log, Phantomhelp flight log, video recording in youtube, Dji Go App flight log video capture in youtube, Blackbox download, everything, so I would really appreciate your help about what happened during this flight.

The flight took place in Sigoldujuglfur, a gorgeous Canyon you can find on Google maps here . As you can see, there's a Power Station at a not very high distance, even though there were no high tension pillars near me. The Spark battery was cold at the beginning (long before the errors), and wind was quite strong (for an AC like the Spark).

Before this flight I had tried to fly the Spark twice in the same place, 10 minutes before, and was getting IMU calibration and lost gps errors, but they happened right at the beginning and then dissapeared. Since, like I said above, I have been getting constant IMU calibration warning which then disappear upon AC restart, and since I got no errors when I started the present flight, I thought everything was ok. Quite honestly, I have no idea how a GPS signal which is perfectly stable, can be suddenly lost in half a second and then have this being followed by all sorts of warnings.

Since I bought the Spark I've already flown it a considerable number of times. In retrospective, it seems that now and then there is always some sort of error messagem that pops up, making the whole experience of flying an exercise of stress. Drones are supposed to make us feel free, but that isn't happening with the Spark. If it weren't for the fact that I use the Spark as a content creation tool, I can't imagine myself using it as a stress-relieving device.

AirData Log - [backcolor=rgba(249, 249, 249, 0.3)]http://app.airdata.com/main?share=zMhVNF

Phantomhelp Log - http://www.phantomhelp.com/logviewer/PKFULDNGKV9KK3865SWY/#

You can view the recorded video of the Spark when things went bad. I guess that it's at 30 seconds that it lost GPS. Notice how fast it quickly drifts, while I standing there watching ...
This is a video capture I made of the iPad screen showing the first two minutes and the last 5 minutes of the flight. As you can see, at the end of the first 2 minutes period there's a sudden GPS loss, which is quickly recovered. The last minutes, well... total chaos!

Flight Controller Log (from my iPad) - https://drive.google.com/open?id=0By0XHzSAT15uTnJKT0tlWXZDU2s

Black Box Download Link (from the Spark) - https://drive.google.com/open?id=0By0XHzSAT15uaXZvbnl0cDlLeUU

Unfortunately I could not download the Data Upload Flight Log through DJI Assistant 2, because the flight files are no longer available! As I had to travel back from Iceland, I had to discharge the Lipo Batteries for the flight, so the quickest way I found out to do it (I could not fly it, it was raining badly on the last day) was to turn on the Spark synced to the RC and iPad, with camera on. Unfortunately it looks like logs for these "flights" were created, overwriting the problematic flight file. Is there any way to retrieve it? (image in attach showing only 5 downloadable logs available)

Thanks a LOT for your help!






Captura.JPG
2017-9-27
Use props
LouisP
Captain
Flight distance : 41952 ft
United States
Offline

I think that it is becoming apparent that Spark still has some software issues that DJI is either refusing to accept or can not seem to fix.  The compass, GPS, and IMU are three totally different systems that can each help the spark maintain stable flight.  Losing one (most often the compass) should not cause the others to suddenly stop working. It is the software that is incapable of deciding what to do when one of its inputs becomes questionable.  It just drops into ATTI mode (basically just keeps the rotors parallel to the ground) and turns total control over the the pilot.   One would think that with a compass error the software should be smart enough to use the IMU to know the direction it was pointing and any movements that the gyros have picked up can keep a close estimate of the current direction.   GPS technology is fairly robust at this point and I really doubt that spark is losing GPS when it does this, especially when it is still indicating multible sat locks.  Still, even a momentary loss of GPS should not mean total failure of the navigation system. GPS might go away for a few seconds and come back.. Again, the gyros in the IMU can predict the position without GPS.  I do hope that some major improvements to the software are being worked on.  I am not one to blame the spark or DJI for everything that happens. But, it is starting to become obvious that the navigation software really needs some improvements.
2017-9-27
Use props
JP Santos
lvl.4
Flight distance : 116929 ft
Portugal
Offline

Did you calibrate your spark while in Iceland?
2017-9-27
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

JP Santos Posted at 2017-9-27 15:02
Did you calibrate your spark while in Iceland?

Yes, right from the first flight. Actually, I started calibrating it first thing every day in Iceland, as there could be some interferences.

Even if this was a compass problem, why would it happen just mid flight? Why would everything else fail after the compass failure?

Since there are so many similar posts about this, I'm starting to think about software failure and/or hardware failure
2017-9-27
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

LouisP Posted at 2017-9-27 14:28
I think that it is becoming apparent that Spark still has some software issues that DJI is either refusing to accept or can not seem to fix.  The compass, GPS, and IMU are three totally different systems that can each help the spark maintain stable flight.  Losing one (most often the compass) should not cause the others to suddenly stop working. It is the software that is incapable of deciding what to do when one of its inputs becomes questionable.  It just drops into ATTI mode (basically just keeps the rotors parallel to the ground) and turns total control over the the pilot.   One would think that with a compass error the software should be smart enough to use the IMU to know the direction it was pointing and any movements that the gyros have picked up can keep a close estimate of the current direction.   GPS technology is fairly robust at this point and I really doubt that spark is losing GPS when it does this, especially when it is still indicating multible sat locks.  Still, even a momentary loss of GPS should not mean total failure of the navigation system. GPS might go away for a few seconds and come back.. Again, the gyros in the IMU can predict the position without GPS.  I do hope that some major improvements to the software are being worked on.  I am not one to blame the spark or DJI for everything that happens. But, it is starting to become obvious that the navigation software really needs some improvements.

Why would a compass error disable GPS?
Orientation (compass data) and position (GPS data) are different but related.
When the Spark is moving and getting position information from the GPS that tells the Spark that it is travelling in a different direction from what it's getting from compass data, The Spark programming can't work with the data conflict.
The solution is to drop data from one source and since the Spark can't fly without a compass but can without GPS, it's the GPS data that gets dropped.
2017-9-27
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

Pardon me, but where is this conclusion coming from?
The Spark's user's manual says that the switch to ATTI mode occurs when neither GPS nor visual sensor data are available. In other words, if you get switched to ATTI mode after a compass error - it means all other systems are virtually dead.
And quite frankly, if GPS data is valid but is dropped - then the algorithm doesn't make any sense. If GPS data is available, at the very least the drone should be able to orient itself back home. point A, current GPS location. Point B, GPS coordinates of the home point. Straight line between them. Boom, you've got direction. Point yourself at that direction and fly home, Mr. Spark! You don't need a compass for that. There's no reason for those flyaways when GPS is available.

If GPS is not available, or is lost and not re-acquired, that's an even worse glitch. There's no reason for a $500 drone to lose GPS and not be able to re-acquire it when 50m in the air with the entire sky, with 15-20 satellites, seen by it.
2017-9-27
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

And just to explain that point in my previous post - you probably ask - how do you "point yourself at that direction" without a compass? easy. start flying, and find out your direction from the change in your coordinates. Correct course accordingly.
Honestly, I don't see why this should be an issue. All it takes is a smarter algorithm.
2017-9-27
Use props
LouisP
Captain
Flight distance : 41952 ft
United States
Offline

djiuser_Kf4iPA3 Posted at 2017-9-27 16:18
And just to explain that point in my previous post - you probably ask - how do you "point yourself at that direction" without a compass? easy. start flying, and find out your direction from the change in your coordinates. Correct course accordingly.
Honestly, I don't see why this should be an issue. All it takes is a smarter algorithm.

I think that you and I are making the same point. Spark has three ways of determining its orientation (not necessarily gps position) Gyros, compass, and GPS.. However, we are seeing time and time again that the loss of any one of them means that the spark goes into ATTI mode.    I agree,, lose the compass and gps says that you are moving, it should be pretty easy to figure out what direction you are moving in.  Lose the GPS, well plenty of drones don't even have gps and fly just fine on gyros alone.  So if you know the position you had when GPS went away, use the IMU to calculate your position, airplanes have been doing that for decades.  Again, this seems like it would be something that could be fixed with a software update. Let's hope that we get one someday.
2017-9-27
Use props
LouisP
Captain
Flight distance : 41952 ft
United States
Offline

Also, notice on the original post, in the second video... his sparks compass was working the whole time.. the little red arrow on the display is moving to show the compass direction of his spark even while he is struggling to bring it back.
2017-9-27
Use props
JeyTea
Second Officer
Flight distance : 399984 ft
Germany
Offline

DJI Elektra Posted at 2017-9-27 19:08
Jose Ramos, please sync your flight record and provide your DJI account to us for further check. You can PM me the account. Thanks for your support.

are your engineers working on that general problem? it is concerning.
2017-9-27
Use props
PedramN
Second Officer
Iran
Offline

djiuser_Kf4iPA3 Posted at 2017-9-27 16:12
Pardon me, but where is this conclusion coming from?
The Spark's user's manual says that the switch to ATTI mode occurs when neither GPS nor visual sensor data are available. In other words, if you get switched to ATTI mode after a compass error - it means all other systems are virtually dead.
And quite frankly, if GPS data is valid but is dropped - then the algorithm doesn't make any sense. If GPS data is available, at the very least the drone should be able to orient itself back home. point A, current GPS location. Point B, GPS coordinates of the home point. Straight line between them. Boom, you've got direction. Point yourself at that direction and fly home, Mr. Spark! You don't need a compass for that. There's no reason for those flyaways when GPS is available.

Hi to all. You are right. In fact such navigation system invented decades ago and used vastly to navigate many unmanned aerial wheicles including missles, airplanes and many more. This called INS or inertial navigation system. With accepatable error margin, INS can navigate every UAV without ANY data from external sources including GPS and tottal independent. See this link https://en.m.wikipedia.org/wiki/Inertial_navigation_system
2017-9-27
Use props
PedramN
Second Officer
Iran
Offline

PedramN Posted at 2017-9-27 21:55
Hi to all. You are right. In fact such navigation system invented decades ago and used vastly to navigate many unmanned aerial wheicles including missles, airplanes and many more. This called INS or inertial navigation system. With accepatable error margin, INS can navigate every UAV without ANY data from external sources including GPS and tottal independent. See this link https://en.m.wikipedia.org/wiki/Inertial_navigation_system

PS: INS reures no external data (including compass) and it is immune to jamming and deception.
2017-9-27
Use props
StarFishUK
lvl.4
Flight distance : 54970 ft
United Kingdom
Offline

I had a VERY similar incident when the Spark went into ATTI mode without warning, mid-flight and began to head for the hills. I also flicked it into Sport mode to attempt to regain control. I felt like I was fighting it back down. Luckily for me the incident only lasted a small number of seconds and I was able to land safely, but I've never looked at the Spark with quite the same level of confidence again. It would be great to hear that this was something they are working on.
2017-9-27
Use props
DJI Elektra
DJI team
Hong Kong
Offline

Jose Ramos Posted at 2017-9-27 15:05
Yes, right from the first flight. Actually, I started calibrating it first thing every day in Iceland, as there could be some interferences.

Even if this was a compass problem, why would it happen just mid flight? Why would everything else fail after the compass failure?

Jose Ramos, please sync your flight record and provide your DJI account to us for further check. You can PM me the account. Thanks for your support.
2017-9-27
Use props
DJI Elektra
DJI team
Hong Kong
Offline

JeyTea Posted at 2017-9-27 20:40
are your engineers working on that general problem? it is concerning.

JeyTea, if you have the same problem, please provide the flight data for us to locate the problem. Our engineers will try their best to solve it. Thanks for support.
2017-9-27
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

DJI Elektra Posted at 2017-9-27 23:18
Jose Ramos, please sync your flight record and provide your DJI account to us for further check. You can PM me the account. Thanks for your support.

Dear DJI, I'll PM you right now.

I've also uploaded the Flight Controller Log (from my iPad) to the original post - https://drive.google.com/open?id=0By0XHzSAT15uTnJKT0tlWXZDU2s

Thanks!
2017-9-28
Use props
El Diablo
First Officer

Greece
Offline

Tip: the author has been banned or deleted automatically shield
2017-9-28
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

El Diablo Posted at 2017-9-28 04:06
Hi Jose, big fan of your work (I guess you are THE Jose Ramos). I bought the Spark for the same reasons as you (I am also a photographer), quite scary and stressful to go flyi knowing that this can happen at any time...

Hi there! Who is THE José Ramos, the landscape photographer from Portugal? If yes, then that's me! ;-)
2017-9-28
Use props
El Diablo
First Officer

Greece
Offline

Tip: the author has been banned or deleted automatically shield
2017-9-28
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

El Diablo Posted at 2017-9-28 04:09
Acho que sim, excelente trabalho!

Eheheheh, thanks a lot! I have a lot of people with their eyes put on me, to see how good is the Spark, so I want everything to go smoothly. :-)
2017-9-28
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

djiuser_Kf4iPA3 Posted at 2017-9-27 16:12
Pardon me, but where is this conclusion coming from?
The Spark's user's manual says that the switch to ATTI mode occurs when neither GPS nor visual sensor data are available. In other words, if you get switched to ATTI mode after a compass error - it means all other systems are virtually dead.
And quite frankly, if GPS data is valid but is dropped - then the algorithm doesn't make any sense. If GPS data is available, at the very least the drone should be able to orient itself back home. point A, current GPS location. Point B, GPS coordinates of the home point. Straight line between them. Boom, you've got direction. Point yourself at that direction and fly home, Mr. Spark! You don't need a compass for that. There's no reason for those flyaways when GPS is available.

No the manual doesn’t say that guide me to the page where it says that visual sensors will cause AC to go into Atti mode?

That is the way all dji drones operate, some have dual compass so won’t go readily into Atti mode. Are you saying that with all the engineers experience in making drones they never taught that it was somehow better to to hold on to gps and dump compass, maybe your in the wrong job. I’ve just explained what happens in the situation when there is conflict between gps and compass and that’s exactly what happens . If compass corrects its self AC will go automatically back to gps mode.

When Aircraft is in Atti mode once in VLOS Pilot should be able to operate craft .
2017-9-28
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

hallmark007 Posted at 2017-9-28 04:18
No the manual doesn’t say that guide me to the page where it says that visual sensors will cause AC to go into Atti mode?

That is the way all dji drones operate, some have dual compass so won’t go readily into Atti mode. Are you saying that with all the engineers experience in making drones they never taught that it was somehow better to to hold on to gps and dump compass, maybe your in the wrong job. I’ve just explained what happens in the situation when there is conflict between gps and compass and that’s exactly what happens . If compass corrects its self AC will go automatically back to gps mode.

Looking at all the data I posted, do you think this was pilot error, a random unpredictable normal consequence or a software/hardware flaw?
2017-9-28
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

Jose Ramos Posted at 2017-9-28 04:24
Looking at all the data I posted, do you think this was pilot error, a random unpredictable normal consequence or a software/hardware flaw?

I don’t think I qualify to give you the answer, I’m not saying there isn’t a problem in your case simply because I don’t know, I’m just explaining what happens when there is conflict between compass and gps.

While I realise you are in Iceland and there is a change in magnetic declination there is a good case for calibrating compass , I’m not sure you should be calibrating every day.

I think this will be up to dji to decide where the fault might lie.
2017-9-28
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

hallmark007 Posted at 2017-9-28 04:39
I don’t think I qualify to give you the answer, I’m not saying there isn’t a problem in your case simply because I don’t know, I’m just explaining what happens when there is conflict between compass and gps.

While I realise you are in Iceland and there is a change in magnetic declination there is a good case for calibrating compass , I’m not sure you should be calibrating every day.

Most places I shot in were 150km apart, so I thought it would be prudent to calibrate the compass.

I also wonder why am I constantly receiving a warning to calibrate IMU before flying, which disappears when I restart the AC...
2017-9-28
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

Jose Ramos Posted at 2017-9-28 04:44
Most places I shot in were 150km apart, so I thought it would be prudent to calibrate the compass.

I also wonder why am I constantly receiving a warning to calibrate IMU before flying, which disappears when I restart the AC...

I don’t know do you have a screenshot of warning, also have you tried calibrating IMU?
2017-9-28
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

hallmark007 Posted at 2017-9-28 04:18
No the manual doesn’t say that guide me to the page where it says that visual sensors will cause AC to go into Atti mode?

That is the way all dji drones operate, some have dual compass so won’t go readily into Atti mode. Are you saying that with all the engineers experience in making drones they never taught that it was somehow better to to hold on to gps and dump compass, maybe your in the wrong job. I’ve just explained what happens in the situation when there is conflict between gps and compass and that’s exactly what happens . If compass corrects its self AC will go automatically back to gps mode.

page 9:

"The aircraft will automatically switch to ATTI mode when the GPS signal is weak
and lighting conditions are too dark for the Vision System, it will not be able to
automatically avoid obstacles and only use its barometer for positioning to control
altitude. Intelligent Flight Modes are not available in ATTI mode. "

and again on page 12:

"3D Sensing and Vision System performance are affected by the type of surface being flown over.
The aircraft will switch from P Mode to ATTI Mode automatically if neither GPS or Vision System
are available. Please be vigilant when operating Spark in the following situations"

Now, please step out of the "defend DJI at all cost, everything is pilot error" box for a minute, and tell me why it makes sense to drop GPS information when it contradicts the compass. GPS is much more reliable than the compass, particularly for a drone flying high where it typically has much better LOS to more satellites than any device at ground level. All the drone has to do is track GPS signal for a little while to find out its direction, correct its course accordingly, and go back home rather than randomly and aimlessly into oblivion. No need for a compass to do that. It may zigzag a little bit at first until it corrects itself, but there's no reason for it to throw away GPS information and fly off in the wrong direction.

It's very easy to say "the pilot should handle it". But if this happens when the drone flies away from you at the edge of your RC range, there's not a lot you can do.
2017-9-28
Use props
Hecate
Second Officer
Canada
Offline

djiuser_Kf4iPA3 Posted at 2017-9-28 05:22
page 9:

"The aircraft will automatically switch to ATTI mode when the GPS signal is weak

Hi,I agree with you 100%, the AC could also adjust the compass value based on GPS data, why drop GPS during a compass error, do just the opposite and tell the compass were north is based on the GPS data.
Just my two cents.
2017-9-28
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

djiuser_Kf4iPA3 Posted at 2017-9-28 05:22
page 9:

"The aircraft will automatically switch to ATTI mode when the GPS signal is weak


Yes the manual says ,gps and vision sensors not one or the other but both.

“Everything is pilot error “

In what post did I say this? And defend dji at all costs.

Regarding a better system you seem to be the expert and maybe you should enlighten us on how you propose to make dji drones safer using your brand of off the cuff ranting without any explanation. Would you not think that dji engineers would be using your theory if it was that simple to engineer.

What you are advocating is something that’s not there, what I’m saying is how it works in all of dji drones that’s fact.

“It’s very easy to say pilot can handle it, but what if it’s on the edge of your RC range”

If it’s on the edge of RC range then it’s almost certainly BVLOS, so how does the pilot handle this .

2017-9-28
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

What I have observed in the flightlogs posted above is that the error messages started at a very early stage of the flight and apparently no effort was made to land the AC immediately but instead the flight was continued. Could the pilot please explain why?
2017-9-28
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

Wachtberger Posted at 2017-9-28 08:52
What I have observed in the flightlogs posted above is that the error messages started at a very early stage of the flight and apparently no effort was made to land the AC immediately but instead the flight was continued. Could the pilot please explain why?

Thanks for checking the flight logs.

Like I previously said, I occasionally got GPS loss messages for a brief seconds, and the AC would quickly regain GPS signal. I was in the last days of my Iceland trip, in one of the most gorgeous places I've ever been to, and only had a single drone, so I opted to continue flight. During this problematic flight I only saw a first error message at around 2 minutes, and the endless sequence of error messages at the end of the flight happened at the same time I lost control of the AC, with great difficulty to bring it back.

Assuming the hypothesis that I decided to fly an AC with an error message which should have made me decide to stop the flight, that shouldn't stop us from investigating if this situation was caused by software/hardware failure.
2017-9-28
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

hallmark007 Posted at 2017-9-28 07:40
Yes the manual says ,gps and vision sensors not one or the other but both.

“Everything is pilot error “

But that's exactly what I said in my original post - "The Spark's user's manual says that the switch to ATTI mode occurs when neither GPS nor visual sensor data are available." That means, in plain English, that both are unavailable. to which you replied "No the manual doesn’t say that".

Many times your reply to people complaining about flyaways was (or hinted in paraphrase, and not in such strong terms, granted) that a decent pilot and/or flying by the rules and/or taking off from where he was supposed to - should have handled it. If you insist I can find some exact quotes.

I'm not a drone engineer, but I do happen to be a radio engineer with more than 25 years of experience. And while my main expertise is cellular, I know my way around the propagation and logic behavior of other transmission systems, GPS included, and electromagnetism in general.
There is no problem for a GPS device, if it's moving in a straight line, to know its direction from a simple vector calculation of the change in its GPS location, and correct its course on the fly. If it has the right algorithm programmed into it, of course. there no sense in not using this data whether you like it or not.

So the one who's ranting without knowledge is you. You keep reciting (in paraphrase) that if DJI did it that way, it must be right and we must accept it without question. The fact is that in a drone that's supposed to be state of the art in its trade, equipped with several navigation systems - when it has a failure in just one of them (and not the most reliable or robust one) it drops the data from ALL of them except the very limited barometer, and flies blindly in a random direction. And that makes sense to you?? you seem to accept it as God given fate.

I'm an engineer, but you don't have to be an engineer to read a little about the technologies your drone is using, rather then keep saying "if they did it, and if it behaves that way, then it must be right and that's it". Then perhaps you'll have some doubts too, maybe you'll see the senselessness of that instead of walking blindly enchanted by the Pied Piper of Hamelin.
2017-9-28
Use props
Wachtberger
Captain
Flight distance : 261509 ft
Germany
Offline

Jose Ramos Posted at 2017-9-28 09:02
Thanks for checking the flight logs.

Like I previously said, I occasionally got GPS loss messages for a brief seconds, and the AC would quickly regain GPS signal. I was in the last days of my Iceland trip, in one of the most gorgeous places I've ever been to, and only had a single drone, so I opted to continue flight. During this problematic flight I only saw a first error message at around 2 minutes, and the endless sequence of error messages at the end of the flight happened at the same time I lost control of the AC, with great difficulty to bring it back.

I fully agree that the resons should be examined and I hope that DJI will be able to provide the answer.
2017-9-28
Use props
LouisP
Captain
Flight distance : 41952 ft
United States
Offline

I think we are overlooking one more thing that seems to be common with many of these cases.  That cryptic error message that pops up on almost all of these " WorkingIMUenconteredexceptionerror" or something like that.  We are talking about what the spark should and should not do if it loses compass or GPS.. but that cryptic error message may be telling something more.  Could it be that the IMU navigation subroutines in the software are crashing for reasons other than loss of signal.   Loss of compass and GPS at the same time might be due to a critical part of the software crashing and taking down both systems at the same time.   Could be the computer is overloaded with work (think about the Apollo 11 landing computer :-) )   Again, if this is the case, it is something DJI must know about and should be working on a fix.  I would not be surprised to see another mandatory software upgrade pushed out in the future. At least I hope so.
2017-9-28
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

That makes much more sense to me than the hypothesis that DJI actually meant it to be that way...
I don't believe they actually MEAN to throw away GPS data when a compass error occurs. Like I said, GPS is pretty robust and there's no sense in deliberately not using it for orientation when a compass error occurs. There's a glitch there that must be fixed. and the exception error, which basically means in computer lingo that a piece of software has crashed and not completed what it was supposed to do, convinces me of that even more.
2017-9-28
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

djiuser_Kf4iPA3 Posted at 2017-9-28 09:13
But that's exactly what I said in my original post - "The Spark's user's manual says that the switch to ATTI mode occurs when neither GPS nor visual sensor data are available." That means, in plain English, that both are unavailable. to which you replied "No the manual doesn’t say that".

Many times your reply to people complaining about flyaways was (or hinted in paraphrase, and not in such strong terms, granted) that a decent pilot and/or flying by the rules and/or taking off from where he was supposed to - should have handled it. If you insist I can find some exact quotes.

When I came to this thread , the Op was looking for answers as to why his AC behaved the way it did, I answered that question , I do the same on all the threads I’ve participated in. So your free to dig up anything you like.
Regarding those using the technology as it is now, well the reality of it is there are literally hundreds of thousands using it with no problems in a proper and safe manner.
Yes I’m sure the technology could be changed but will this make it better or will this lead to other problems greater than we have now.
What do we know. Well we know that dji are innovators of great technology and by a long way make the best consumer drones in the world.
But you quip about the pied piper yes I am a fan of dji’s products , I don’t see anything wrong with that. I also accept for my part responsibility for how and where I use them and from what I read this seems to be the crux of the problem with you.
Far from pied piper, what we have here is “THE BOY WHO CRIED WOLF” it’s clear from this thread that the OP managed in adversity to control and bring his AC to a safe landing.
But it seems from your thread you weren’t capable of doing the same so CRY WOLF it’s the fault of the technology it’s dji’s fault it’s my fault in fact it’s everyone’s fault except THE BOY WHO CRIED WOLF.

The easiest way is to blame everyone and everything else for your own short comings. It’s just a pity you had to hijack OP’s thread to do it.
I’ve no more to say good bye.
2017-9-28
Use props
Jose Ramos
Second Officer
Flight distance : 72825 ft
Portugal
Offline

hallmark007 Posted at 2017-9-28 10:19
When I came to this thread , the Op was looking for answers as to why his AC behaved the way it did, I answered that question , I do the same on all the threads I’ve participated in. So your free to dig up anything you like.
Regarding those using the technology as it is now, well the reality of it is there are literally hundreds of thousands using it with no problems in a proper and safe manner.
Yes I’m sure the technology could be changed but will this make it better or will this lead to other problems greater than we have now.

Yes, I think it's best to end your discussion, so that we can focus on my issue. ;-) I managed to return my Spark home manually, and yet I still think there's a high possibility that something went wrong which is not related with environmental issues or pilot error. I think it's important to understand that people get a bit dissatisfied, when they spend 800USD on a drone which is advertised as being almost fool proof, and then the whole experience becomes a constant source of stress.

Many people here complained about the Sparks falling from the sky, and a misterious mandatory upgrade was most probably related with that, when there were so many users saying it was user error. It's true that most Spark users that resort to this forum are probably doing it because they had a problem with their AC, but it's very important to take into consideration that the number of complaints here about a specific issue is much LESS than the actual number of complaints. Most people with errors like these will sent a support ticket, call Dji, activate dji refresh, write on facebook or go to the store where they bought their drone. Forum culture has been quite abandoned because of social media, so I think we should pay a LOT of attention when there's a series of people here complaining about a similar serious problem with the Spark, during a specific chronological time, as it may be a solid sign of a more widespread problem.  

By the way, I liked the latest thoughts about the strange IMU error, as I always found that error quite cryptic....
2017-9-28
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

hallmark007 Posted at 2017-9-28 10:19
When I came to this thread , the Op was looking for answers as to why his AC behaved the way it did, I answered that question , I do the same on all the threads I’ve participated in. So your free to dig up anything you like.
Regarding those using the technology as it is now, well the reality of it is there are literally hundreds of thousands using it with no problems in a proper and safe manner.
Yes I’m sure the technology could be changed but will this make it better or will this lead to other problems greater than we have now.

Actually personally I haven't had flyaways, I haven't flown enough yet ... but I did have intermittent switches to ATTI mode even in the small amout of milage I've flown (without compass error though). In my cases GPS was re-acquired. That can happen and I don't complain about that.
And I'm a fan of DJI too, mind you - I bought a product of theirs, didn't I?
But that doesn't mean I have to accept everything as sensible, especially considering I know the technologies involved at least from a radio perspective, and I know what they're capable of. Those flyaway cases could have been prevented if the capabilities of the technologies, especially GPS, would have been used for recovery. And according to you, the data is there; so it makes even less sense to deliberately not use it.
DJI engineers are not fools, of course. I agree with you it's an innovative company and a pioneer in its field. That is exactly why I DON'T believe they have deliberately done this.
But I see that even when DJI's software itself declares an exception error, it doesn't convince you there's a problem. Rather, the loss of ALL NAVIGATION from just a compass error seems to you to be fate, something that us simple mortals have to accept without question, because the gods of DJI work in mysterious ways.
2017-9-28
Use props
hallmark007
Captain
Flight distance : 9812789 ft
  • >>>
Ireland
Offline

Jose Ramos Posted at 2017-9-28 10:38
Yes, I think it's best to end your discussion, so that we can focus on my issue. ;-) I managed to return my Spark home manually, and yet I still think there's a high possibility that something went wrong which is not related with environmental issues or pilot error. I think it's important to understand that people get a bit dissatisfied, when they spend 800USD on a drone which is advertised as being almost fool proof, and then the whole experience becomes a constant source of stress.

Many people here complained about the Sparks falling from the sky, and a misterious mandatory upgrade was most probably related with that, when there were so many users saying it was user error. It's true that most Spark users that resort to this forum are probably doing it because they had a problem with their AC, but it's very important to take into consideration that the number of complaints here about a specific issue is much LESS than the actual number of complaints. Most people with errors like these will sent a support ticket, call Dji, activate dji refresh, write on facebook or go to the store where they bought their drone. Forum culture has been quite abandoned because of social media, so I think we should pay a LOT of attention when there's a series of people here complaining about a similar serious problem with the Spark, during a specific chronological time, as it may be a solid sign of a more widespread problem.  

I think the only way to find out what happened is to get dji to take a look at your dat file , they should be able to give you some more information.
2017-9-28
Use props
Pedro-BR
lvl.1
Flight distance : 36375 ft
Brazil
Offline

LouisP Posted at 2017-9-28 10:04
I think we are overlooking one more thing that seems to be common with many of these cases.  That cryptic error message that pops up on almost all of these " WorkingIMUenconteredexceptionerror" or something like that.  We are talking about what the spark should and should not do if it loses compass or GPS.. but that cryptic error message may be telling something more.  Could it be that the IMU navigation subroutines in the software are crashing for reasons other than loss of signal.   Loss of compass and GPS at the same time might be due to a critical part of the software crashing and taking down both systems at the same time.   Could be the computer is overloaded with work (think about the Apollo 11 landing computer :-) )   Again, if this is the case, it is something DJI must know about and should be working on a fix.  I would not be surprised to see another mandatory software upgrade pushed out in the future. At least I hope so.

I really look forward for this...

To open the DJI GO and receive the firmware update popup!
2017-9-28
Use props
djiuser_Kf4iPA3
First Officer
Flight distance : 510968 ft
Israel
Offline

Jose Ramos Posted at 2017-9-28 10:38
Yes, I think it's best to end your discussion, so that we can focus on my issue. ;-) I managed to return my Spark home manually, and yet I still think there's a high possibility that something went wrong which is not related with environmental issues or pilot error. I think it's important to understand that people get a bit dissatisfied, when they spend 800USD on a drone which is advertised as being almost fool proof, and then the whole experience becomes a constant source of stress.

Many people here complained about the Sparks falling from the sky, and a misterious mandatory upgrade was most probably related with that, when there were so many users saying it was user error. It's true that most Spark users that resort to this forum are probably doing it because they had a problem with their AC, but it's very important to take into consideration that the number of complaints here about a specific issue is much LESS than the actual number of complaints. Most people with errors like these will sent a support ticket, call Dji, activate dji refresh, write on facebook or go to the store where they bought their drone. Forum culture has been quite abandoned because of social media, so I think we should pay a LOT of attention when there's a series of people here complaining about a similar serious problem with the Spark, during a specific chronological time, as it may be a solid sign of a more widespread problem.  

Well said,
The first step to solving this issue is acknowledging there is one... like the falling Sparks of before.
While we're at it, does DJI issue release notes to their firmware updates?

By the way, I love your country one of my favorite countries in the world! I wish I had the Spark when I was there 4 years ago...
2017-9-28
Use props
12Next >
Advanced
You need to log in before you can reply Login | Register now

Credit Rules