hallmark007 Posted at 2017-10-10 07:46
I fully agree, and while I’ve explained what happens in this situation, it doesn’t mean loss of aircraft and there is a redundancy for this just as there is for loss of gps in fact they are the same.
If there is a firmware causing compass problems then this should be fixed although I haven’t seen any proof that this is the case, I do believe as people gain more experience on how and where to fly these Aircraft we will see a lot less of what we are seeing for what is a very new aircraft and a lot of first time users.

Firstly, keep in mind that I am someone that is a very strong proponent of personal responsibility and have made a video about this in relation to the Spark as well as many posts on this forum of the same.

I see LOTS of instances where drone pilots make terrible decisions that end up leading to bad results.  One of the videos I made discusses breaking the chain of events that lead to a potential disaster.  It is very rare that it is a single thing that leads to a lost drone.  

In my opinion based on what I've seen in logs from users posted in this forum, one of the events in that chain quite frequently ends up being the compass error which results in AC going into ATTI mode.  As a pilot, choosing to fly in a better location may have avoided the situation from the beginning.  As a pilot they can also be better prepared for this situation so that if/when this happens, they are able to adequately prepared fly their AC to safety.  I think this is what you are advocating and I strongly agree.

However, in my opinion, that shouldn't be the last failsafe point.  Due to the audience the Spark is being marketed to (new users with no experience or training)  as well as the potential for other links in the chain to go wrong (i.e. connection lost), I think there should be an additional software failsafe to break that chain.  I think if you read my proposal post #73 of this thread, you may find yourself agreeing with what I'm talking about.

Let's not try to conflate different error sources and assume the solution to them is the same.  I think I agree with you that with either a Compass or GPS error, the intermediary solution is to hand control to the pilot in ATTI mode and let them try to save themselves.  However, in the case of a Compass error, if they aren't able to save themselves and the drone loses connection with the controlling device, why not initiate an RTH procedure that uses the resources available on the AC to get it back home?

In the case of a GPS error, you are right to say that it should switch to ATTI mode.  In this case, I don't have a proposed solution if the AC then also loses connection to the controlling device.  But just because a good solution may not be available for this particular case doesn't mean that a software solution shouldn't be provided for the Compass error use case.

In my opinion as a pilot, I should do everything I can to prevent problems and be as best prepared as I can to handle a situation where errors happen.  However, I strongly believe that DJI should also be doing everything they can from their side to implement failsafe features from their side to prevent lost drones as well.

To think that there may not be some sort of a problem with the Spark (either software or hardware) that contributes to a portion of the Compass error ATTI mode flights might be a bit of a stretch.  Keep in mind that all of the other DJI drones above the Spark (including Mavic Pro, Phantom, etc) have redundant systems (2 compass, 2 gyros, etc, etc).  The Spark does NOT have this same redundancy built into it.  So DJI has already added a link to the chain by deciding not to make the Spark with redundant systems.

The Mavic Pro had a documented issue with electromagnetic interference when flying fast that was causing compass problems.  They fixed the issue by failing over to the other compass mid-flight.  So it is not hard to imagine some sort of software or hardware issue that POTENTIALLY could be involved with a percentage of these Spark errors that occur.   See this thread about Mavic issue:  https://forum.dji.com/thread-76005-1-1.html
  

HereForTheBeer Posted at 2017-10-10 08:00
i stand by i think dji is wrong in ignoring perfectly usable GPS data, theta data.  its wasteful, it should be augmented differently not ignored because the compass is freaking out..  the answer lies in the middle, i dont know what is best but the current solution i think most can agree on is both wasteful and problematic..  

like here is one simple example of how it could function.. its overly simplified but its part of data augmentation idea..  say compass is derped out, and in RTH mode and drone fly up and away, instead of continuing the wrong way it could see that its not going the right way by coordinates its crossing not being what is expected, it could then virtually attenuate the compass data rotating the drone around and then trying again coming back where it believes it should using some simple logic .. knowing which coordinates are which way then simply back tracking its incorrect move and correcting it..

I agree, if the GPS works, use it to its fullest.  

I also like the idea of hovering and giving the user a slider before going full ATTI mode.  Just this simple step would really allow someone to get their wits about themselves.

djiuser_Kf4iPA3 Posted at 2017-10-10 08:42
No, the problem is losing DECENT COMPASS. Then it becomes a BAD COMPASS... and That's what he wrote
that's what starts all the trouble - losing DECENT COMPASS and HAVING BAD COMPASS.

Again your just being a ridiculous narcissist, if you fly to far with not enough battery to get home you will lose your Aircraft.
If you don’t look after your Aircraft or fly in unsuitable environments ie your compass or flying in areas where you may experience magnetic interference or loss of gps then you risk the chances of loosing your Aircraft.
If you choose to fly in safe environments and look after your aircraft then you won’t have the problems your talking about.
What your looking for is a drone for dummies that’s not what these drones are. Maybe you should concentrate on obstacle avoidance for formula 1 racing.

As I said if there’s a problem with firmware or hardware then this should be fixed, and apart from your tech spec you still have shown no proof that your system will work, if you have true belief in your system then let’s see it working but until then, look after your aircraft fly in suitable environments and you won’t have a problem with compass.

One thing I agree with you though:
If the compass has bad firmware (and sure it does... an angular error of 350 deg? come on!!) then it has to fixed. It had to be fixed yesterday. Regardless of anything we're discussing here...
let any procedure for compass loss deal with REAL compass loss. Due to REAL reasons. Not loss due to such bugs.

Hecate Posted at 2017-9-30 10:34
Yes you are correct (note: 2nd picture you have a typo the 10 should be 16). No error at 20 & 19 but having errors at 10 & 16 is interesting. However the one thing that stands out is that the MagYaw line  should disappear at the top of the chart while appearing on the bottom but it does not always do that, the MagYaw lines are crossing the screen. Visualizing that it is rotating fully left and right without crossing the transition line, and it looks like it does that at the transition line only. So what this chart tells me is that system is measuring a yaw error of 344 degrees while in reality both points are only 10 or 16 degrees apart.
I have put the chart together in a different way to show what is happening.

And that may be the problem.

I am curious - in which module its calculated.
Anyone tried to replicate that issue?
It should be replicable by flying directly to the south, to keep a nose around 180 degrees..

In fact exit from error state should be a simple yaw rotation where there will not be wrong calculation difference.

"In the case of a GPS error, you are right to say that it should switch to ATTI mode.  In this case, I don't have a proposed solution if the AC then also loses connection to the controlling device.  But just because a good solution may not be available for this particular case doesn't mean that a software solution shouldn't be provided for the Compass error use case."


After all, on a catastrophic absence of resources (GPS and/or Compass) I think the better is to HOVER IN PLACE (expecting/waiting things to become ok again) an on 10% battery, start autolanding – The WORST scenario is to drift with the wind.

But… how would you HOVER with no GPS?

        1 – Vision System
                  1.1        – But is works up to 7 meters only

2 – Why don´t face the camera down and autotrack any object on the groud?]


Whell, just and IDEA.

What I know its that there is a lot of options / failsafes that would be better to use than switch to ATTI and drift to the wind. Don´t know why those actions are not implemented.

hallmark007 Posted at 2017-10-10 09:09
Again your just being a ridiculous narcissist, if you fly to far with not enough battery to get home you will lose your Aircraft.
If you don’t look after your Aircraft or fly in unsuitable environments ie your compass or flying in areas where you may experience magnetic interference or loss of gps then you risk the chances of loosing your Aircraft.
If you choose to fly in safe environments and look after your aircraft then you won’t have the problems your talking about.

Once again you don't read. If you don't like the battery situation, nobody is stopping you from canceling the procedure and taking over the craft manually. But you want to FORCE everyone to fly manually, even though the compass can be backed up by GPS in case of failure. No matter what. Even if they are not that good or experienced or confident, and even if the battery has 80%, 85%, 90% or 99% left.

I mean, why do you condone the smart navigational systems and the RTH procedure? for redundancy, you say; Good. How about a better flying experience with more confidence, even for novices; How about reducing the possibility of human error during an emergency; How about that?
Your entire position in this argument goes fully against your own rationale about redundancy. One nav system fails = all nav systems fail, no matter what, even in cases in which the remaining working techs can cover up for the bad one. Compass fails? You must fly manual, you SOBs!

BTW, have you noticed another guy has proposed a GPS-only-RTH procedure in case of compass failure? That must not be allowed! Go ahead and demand that he build and propgram a drone by himself to prove his point... or just simply go call him names. Why only me?
But that's pretty much all you have left to offer in this discussion.

WOW ... this is impressive analysis.  This is way over my head

Kloo Gee Posted at 2017-10-10 07:17
This is a fascinating discussion and something that I've had on my mind for quite some time when looking at the errors I see in all the logs I see on the forum.  I'm what I like to call a technical idiot!  I know just enough to think I know how everything works, but just not quite enough to know I am probably missing a key piece of knowledge that may make or break a theory.  So here goes my thoughts:

First off, DJI needs to solve the root cause of compass error whether it is a software issue or a hardware issue.  If they solve this, then most (but not all) of the reason for this discussion goes away.  

Hey man.
Like I said, the procedure you offered is the best post I read here in a while. Actually, it's probably the best post I read ever since I joined this forum. and if you give up obstacle avoidance, it's actually better than the one I had suggested - faster and more battery-efficient. Salutes to you!

I have one question: in another post you suggested not to initiate this procedure immediately upon compass failure, but let the pilot fly ATTI mode first; and if this goes wrong, only then initiate the procedure. I hope I understood you correctly; if not, please correct me.
My question is - why not initiate the procedure immediately, and let the users choose whether and when they want to stop it (the same way they can stop regular RTH) and move to ATTI mode at will? Why force them to start with ATTI?

Thanks man, and once more, well done.

Kloo Gee Posted at 2017-10-10 09:03
Firstly, keep in mind that I am someone that is a very strong proponent of personal responsibility and have made a video about this in relation to the Spark as well as many posts on this forum of the same.

I see LOTS of instances where drone pilots make terrible decisions that end up leading to bad results.  One of the videos I made discusses breaking the chain of events that lead to a potential disaster.  It is very rare that it is a single thing that leads to a lost drone.  

I think we should start with what level we are at here, I feel we can all get carried away with ourselves, this is a $500 Drone witch is packed with the technology it deserves for the price, that’s the starting point, so yes extra redundancy features will cost money whether its research and development costs or whether it’s adding extra hardware, so now we are into a different aircraft.

Sometimes people want cake and eat it, yes Mavic has 2 compass and 2 IMU and yes it would be great if spark had the same but I suspect it would become more expensive and size would need to increase.
I’m also pretty sure that safety is very high on dji’s agenda and if a  better failsafe for compass problem issue can be implemented it should and I’m almost certain that unless it has a much higher cost and it works it will almost certainly be implemented.

It’s not that long ago that P4 was released a great Aircraft with only 1 compass and 1 Imu it didn’t take long for dji to release p4p which they added 2 compass and 2 Imu for little extra cost so I don’t think dji are lacking in providing redundancies and we see it coming on board more and more faster and faster.

Yes if there is something wrong with firmware dji should fix it, but we all know that flying these aircraft is always a risk, risk for a lot is part of the buzz. I think if we could fly these from our armchair they wouldn’t have half the appeal.

Loosing gps can also be solved with bigger and more expensive GPS receivers or another GPS receiver on the outside of Aircraft , but again we are looking at a different aircraft and a lot more expensive.

I think for its size and price spark has by far the best technology of any similar aircraft on the market, but we will lose the run of ourselves if we think we can have all the safety features of a matrice 200 then we are living in cuckoo land,

I do believe as time goes on we will see improvements both in safety features and technology in the spark, and just like the P4Pro had a higher price tag so will the spark.
I tend to take a pragmatic view when someone pops up with a Utopian idea which will cost no money has never been tested I.e. in the air and can say without any doubts that it will work with what is already there , can dismiss everybody else who tries to explain why it may not work. Answer to this should be grow a pair go try it out see if it works , I’ve seen so many failed ideas over the years simply because they never got beyond white paper. Call me a pragmatist or realists but I’m afraid I would need to see it in action as I’m sure most would before I accept it’s a better and safer way.

djiuser_Kf4iPA3 Posted at 2017-10-10 10:02
Once again you don't read. If you don't like the battery situation, nobody is stopping you from canceling the procedure and taking over the craft manually. But you want to FORCE everyone to fly manually, even though the compass can be backed up by GPS in case of failure. No matter what. Even if they are not that good or experienced or confident, and even if the battery has 80%, 85%, 90% or 99% left.

I mean, why do you condone the smart navigational systems and the RTH procedure? for redundancy, you say; Good. How about a better flying experience with more confidence, even for novices; How about reducing the possibility of human error during an emergency; How about that?

The difference between you and I is you know nothing about flying, you keep harping on about RTH , when your in an emergency situation the first thing you must do is to land your aircraft safely as soon as possible , no matter what aircraft manned or unmanned that’s the procedure not RTH . Not sure if you are aware once you hit RTH you now have an aircraft no longer under the control of the pilot. That’s fact don’t be teaching anything else . So while it’s a form of redundancy it should never be your first option. You should learn how to fly first . Or just sit at home while your spark flys around the garden.

djiuser_Kf4iPA3 Posted at 2017-10-10 10:17
Hey man.
Like I said, the procedure you offered is the best post I read here in a while. Actually, it's probably the best post I read ever since I joined this forum. and if you give up obstacle avoidance, it's actually better than the one I had suggested - faster and more battery-efficient. Salutes to you!

Thanks for the feedback.  In my main post with the proposal, this is specifically what I said:
“Philosophically, the question is when to initiate this Compass Error Mode RTH procedure?  Do you do it at the point today where the Spark goes into ATTI mode?  Or do you do it at the point where the remote control loses connection AND you are in ATTI mode from a Compass error?  I personally would recommend the latter, but might suggest an option in the menus to allow a user to choose.”

Honestly I don’t feel too strongly with either option. That is why I think it would be better to give an option for the user to decide which behavior they want to happen. Then if someone feels strongly either way, they can choose for themselves.  For myself I guess I kind of lean towards letting the user actually try to fly the thing.  But from a safety standpoint, it might be safer to just initiate the RTH immediately.  However, when in doubt, let the user choose.  That covers the safety for a new user and the customizability for a more advanced user.

Hallmark:
I don't teach anything there, it's the Spark's manual that teaches that. According to the manual, any RTH procedure (except critical battery level) can be stopped at any time you have RC connection to the AC, and manual control can be resumed. Smart RTH, low battery RTH, even failsafe RTH once connection to the RC has been resumed - they can all be stopped at will. That's what the manual says.
If you have an argument with that, take it up with DJI.

So I'm only suggesting to have what other RTH modes have - the OPTION to resume manual control at will. as opposed to FORCING it.
If you want to fly the AC manually right from the start of compass loss, my procedure doesn't deny you that option. But you want to force it on everyone else.

Kloo Gee Posted at 2017-10-10 10:52
Thanks for the feedback.  In my main post with the proposal, this is specifically what I said:
“Philosophically, the question is when to initiate this Compass Error Mode RTH procedure?  Do you do it at the point today where the Spark goes into ATTI mode?  Or do you do it at the point where the remote control loses connection AND you are in ATTI mode from a Compass error?  I personally would recommend the latter, but might suggest an option in the menus to allow a user to choose.”

Fair enough. To each his own - I get the rationale. That makes sense.
As long as you have the procedure in place, rather than force ATTI on everyone when it can be avoided... and if the Charles Lindberghs of the world would like to cancel it and use ATTI - they can knock themselves out.

Well done on that, too.

Adriano Araujo Posted at 2017-10-10 09:53
"In the case of a GPS error, you are right to say that it should switch to ATTI mode.  In this case, I don't have a proposed solution if the AC then also loses connection to the controlling device.  But just because a good solution may not be available for this particular case doesn't mean that a software solution shouldn't be provided for the Compass error use case."

I agree, if there is a functioning system that can assist, then they should use it.  I would hope they would use VPS if close enough to the ground, but I’ve not seen whether they do that or not. As for using the regular camera for tracking, I’m not sure they’ve done anything with it so far on the Spark, so I *think* that one may be a bit harder to implement and utilize.

hallmark007 Posted at 2017-10-10 10:20
I think we should start with what level we are at here, I feel we can all get carried away with ourselves, this is a $500 Drone witch is packed with the technology it deserves for the price, that’s the starting point, so yes extra redundancy features will cost money whether its research and development costs or whether it’s adding extra hardware, so now we are into a different aircraft.

Sometimes people want cake and eat it, yes Mavic has 2 compass and 2 IMU and yes it would be great if spark had the same but I suspect it would become more expensive and size would need to increase.

Totally agree. It is easy to say 'add this and that to make it perfect' but lets pull back a bit and address what product we have in hand... This is a £500 drone that has one compass, one IMU, is the size of a mobile phone and uses WiFi transmission with it's major USP being the fact that it's flight can be controlled with palm movements... That is why it is £500 (a twelth of the price of the I2 X5S). From a commercial perspective, it is basic in order to be cost efficient for the manufacturer - this is not a drone designed to carry buckets of water, be flown miles from the operator, hundreds of metres AGL - look at it - it's an entry level mini drone aimed at those in their mid teens and above. I love the Spark as it is so inoffensive and innocuous and great fun to fly and provides excellent footage when flown in VLOS. The problems of signal loss, IMU and compass errors occur when the drone is pushed beyond its safe limits and the pilot forgets that they are not flying a £1000 Mavic or £1600 P4P or £6000 I2 with OcuSynch and Lightbridge technology and a list of redundancies - you are flying a Spark - know it's limitations and fly it as it is designed to be flown - keep it away from built up areas, electromagnetic interference and ensure you can see it with your naked eye at all times and accept the limitations of what is in your control. Abuse these limitations and you increase the risk of never seeing your Spark again and YOU will be responsible. I don't think DJI are going to add redundancies to this mini drone that will enable it to be flown like a Mavic - once again, it's £500 for a reason.

Kloo Gee Posted at 2017-10-10 11:01
I agree, if there is a functioning system that can assist, then they should use it.  I would hope they would use VPS if close enough to the ground, but I’ve not seen whether they do that or not. As for using the regular camera for tracking, I’m not sure they’ve done anything with it so far on the Spark, so I *think* that one may be a bit harder to implement and utilize.

When flying in ATTI the AC will not switch to OPTI mode when landing manually, however; when selecting the auto landing the AC will switch to OPTI and use its VPS to land.

ACW:
I don't expect there to be the redundancy level of bigger and more expensive drones such as having two of each system. That's not realistic, I totally agree, and I never advocated that. But I would like to have the drone implement the potential redundancies that its existing hardware allows, as in if system A fails and system B has the capability to back it up - have that option. I don't think it's an unreasonable thing to ask.

I would also not go as far as saying that every fault of any system is because of the pilot. You say people are too harsh on DJI sometimes? maybe they are, sometimes. granted. But claiming that all faults in all systems are because of pilots overstretching it - that's extreme the other way.
Just as a quick example, something like the compass failure presented at the very beginning of this thread - presenting an error of 350 degrees instead of 10, or 344 instead of 16 or anything of that sort - I don't see how pilot performance could have caused that.

djiuser_Kf4iPA3 Posted at 2017-10-10 11:35
I don't expect there to be the redundancy level of bigger and more expensive drones such as having two of each system. That's not realistic, I totally agree, and I never advocated that. But I would like to have the drone implement the potential redundancies that its existing hardware allows, as in if system A fails and system B has the capability to back it up - have that option. I don't think it's an unreasonable thing to ask.

I would also not go as far as saying that every fault of any system is because of the pilot. You say people are too harsh on DJI sometimes? maybe they are, sometimes. granted. But claiming that all faults in all systems are because of pilots overstretching it - that's extreme the other way.

There will always be issues with any device or drone you purchase - I've seen videos on YT of a $10,000 I2 fall from the sky because the new prop locks stuck mid flight - all tech will have issues. My point is that we need to accept that these issues happen and even more so when you are talking about entry level - you get what you pay for and need to adjust to the product in hand. I fly all of my drones very differently in line with their limitations. There are probably hundreds of Sparks with minor faults, handfuls with major ones and thousands without any issues. I took my Spark out yesterday and no matter what I did the gesture control to start video recording wouldn't work - oh well! I just switched to manual flight, pressed record, switched to gesture and guess what - it recorded the rest of my active track in gesture control. It was probably due to the degree I raised my arm or something stupid and will probably work next time but I accept this is not a perfect drone and I didn't put a thread on here "why doesn't my gesture record work anymore". If DJI can implement firmware to improve flight safety then great and more of it but until then we all need to appreciate the limitations, adjust to those limitations and stop chasing our tails on here and pretending our £500 drone is faultless. I've never had an IMU failure or lost GPS though with my Spark and I for one don't believe in luck...

djiuser_Kf4iPA3 Posted at 2017-10-10 10:52
Hallmark:
I don't teach anything there, it's the Spark's manual that teaches that. According to the manual, any RTH procedure (except critical battery level) can be stopped at any time you have RC connection to the AC, and manual control can be resumed. Smart RTH, low battery RTH, even failsafe RTH once connection to the RC has been resumed - they can all be stopped at will. That's what the manual says.
If you have an argument with that, take it up with DJI.

Again the man who thinks he reads everything. I will repeat once you push RTH you now have an Aircraft no longer under the control of pilot you should also be aware that’s why you get continuous beeping.

And forget about your RTH if that’s your first option in an emergency as I said you know nothing about the safety of flying. What your looking for is someone to hold your hand while your flying.

And no you won’t learn this in the manual, you will learn this in flight school.

Kloo Gee Posted at 2017-10-10 09:07
I agree, if the GPS works, use it to its fullest.  

I also like the idea of hovering and giving the user a slider before going full ATTI mode.  Just this simple step would really allow someone to get their wits about themselves.

Just a quick remark there - hovering is not a viable option.
Without a compass, If the craft hasn't moved on its own initiative, it has no idea where it's facing. Therefore, if it tries to hover and drifts - even though it would know the azimuth of drift from GPS, it wouldn't know which way to go to correct itself.

Kloo Gee Posted at 2017-10-10 11:01
I agree, if there is a functioning system that can assist, then they should use it.  I would hope they would use VPS if close enough to the ground, but I’ve not seen whether they do that or not. As for using the regular camera for tracking, I’m not sure they’ve done anything with it so far on the Spark, so I *think* that one may be a bit harder to implement and utilize.

All obstacle avoidance is off in Atti mode .

hallmark007 Posted at 2017-10-10 10:20
I think we should start with what level we are at here, I feel we can all get carried away with ourselves, this is a $500 Drone witch is packed with the technology it deserves for the price, that’s the starting point, so yes extra redundancy features will cost money whether its research and development costs or whether it’s adding extra hardware, so now we are into a different aircraft.

Sometimes people want cake and eat it, yes Mavic has 2 compass and 2 IMU and yes it would be great if spark had the same but I suspect it would become more expensive and size would need to increase.

I think we all understand and can agree that we are talking about DJI's entry level drone and that for the price point, it is a technological marvel!  I think we can also all agree that the given hardware specifications of the Spark are likely largely in consideration of that price point.

I don't think anybody here is making any suggestions to change that hardware because those limitations are understood.  I think from my perspective, we are just suggesting DJI to utilize the given hardware they do have implemented to its fullest extent.

I also don't think that anybody here is trying to claim any sort of Utopian free solutions.

From a  technology company's perspective, there is a cost associated with implementing additional features as you mention.  However, the part nobody wants to mention is that for a technology company, there is quite often also a cost for NOT implementing a new feature.  

In this case, I think it is quite straightforward for a company like DJI to calculate the number of engineering hours to design, implement, and test something similar to what I proposed in post #73.

For choosing to NOT implement the feature, the costs are in number of support tickets opened, number of hours the support team spends working those issues.  The number of hours analyzing the blackbox data and flight records for those tickets as well as the cost of replacing those devices where its determined to be a warranty issue.  The harder costs to calculate are lost sales to new users due to bad PR, lost sales upsells to existing customers who lost confidence in company, etc, etc.

I know I'm not saying anything anybody here doesn't already know, but sometimes it is easy to only look at one side of the cost equation and forget or ignore the other side of the cost equation.

As for whether we are talking only white paper theories that are crap or real features that could be implemented, only DJI really knows the answer to that.  You and I both know that is impossible for any member of this forum to prove their idea since DJI does not provide an open source environment for us to be able to implement our ideas.

However, I do feel strongly if my stated assumptions are correct, my kludgy procedure could very easily be implemented and someone far smarter than myself could implement something far more elegant.

Hecate Posted at 2017-10-10 11:32
When flying in ATTI the AC will not switch to OPTI mode when landing manually, however; when selecting the auto landing the AC will switch to OPTI and use its VPS to land.

I think you are very likely correct.  I guess the question that should be asked is why that is the case?  If you have a functioning system that can assist with guidance, why not use it?

ACW Posted at 2017-10-10 11:08
Totally agree. It is easy to say 'add this and that to make it perfect' but lets pull back a bit and address what product we have in hand... This is a £500 drone that has one compass, one IMU, is the size of a mobile phone and uses WiFi transmission with it's major USP being the fact that it's flight can be controlled with palm movements... That is why it is £500 (a twelth of the price of the I2 X5S). From a commercial perspective, it is basic in order to be cost efficient for the manufacturer - this is not a drone designed to carry buckets of water, be flown miles from the operator, hundreds of metres AGL - look at it - it's an entry level mini drone aimed at those in their mid teens and above. I love the Spark as it is so inoffensive and innocuous and great fun to fly and provides excellent footage when flown in VLOS. The problems of signal loss, IMU and compass errors occur when the drone is pushed beyond its safe limits and the pilot forgets that they are not flying a £1000 Mavic or £1600 P4P or £6000 I2 with OcuSynch and Lightbridge technology and a list of redundancies - you are flying a Spark - know it's limitations and fly it as it is designed to be flown - keep it away from built up areas, electromagnetic interference and ensure you can see it with your naked eye at all times and accept the limitations of what is in your control. Abuse these limitations and you increase the risk of never seeing your Spark again and YOU will be responsible. I don't think DJI are going to add redundancies to this mini drone that will enable it to be flown like a Mavic - once again, it's £500 for a reason.

I would rather learn how to fly and take care of my spark correctly than have to pay for extra for stuff it doesn’t need as you say it’s not designed to fly miles doesn’t have the battery for it either but some around think it’s an inspire2 with the new Zenmuse x7 but they will only pay €500 quid for it, as far as my experience with it it’s a great little machine takes great photos easy launched easy landed plenty of safety features .
The first thing for a safe aircraft is a safe pilot .

hallmark007 Posted at 2017-10-10 11:49
Again the man who thinks he reads everything. I will repeat once you push RTH you now have an Aircraft no longer under the control of pilot you should also be aware that’s why you get continuous beeping.

And forget about your RTH if that’s your first option in an emergency as I said you know nothing about the safety of flying. What your looking for is someone to hold your hand while your flying.

Like I said, take that argument up with DJI, because they're the ones saying you can stop any RTH procedure at will. Not me.

Here's what they say:
Smart RTH: "The user can also immediately exit Smart RTH by pressing the Stop icon in DJI GO4.. Pressing and holding the the RTH button on the remote controller can also initiate Smart RTH. Press it again to exit."
Low battery RTH: "The user can cancel the RTH procedure by pressing the RTH button or Pause Flight button on the remote controller".
Failsafe RTH: "The user may cancel Failsafe RTH and regain control when the wireless signal connection is reestablished."

The only RTH mode which can't be cancelled is critical battery level, which means immediate landing on the spot.
Their words, not mine. You're barking up the wrong tree. Take that argument up with them.
I'm only suggesting employing the same logic they do - allow cancellation of the procedure at will, at the user's choice. Emphasis on "choice".

As for the emergency - if you feel you must land immediately, take over and land immediately. You can also have a setting in which below a certain battery level, or a combination of battery level and distance from home, manual control is mandated, fine. But you want to force ATTI even if you have 100% battery and at any distance from home... Not everyone is as Charles Lindbergh as you are. Just like they have RTH in their service rather than having to fly home manually, they can use my procedure (or that of Kloo which is even better). Same rationale.

hallmark007 Posted at 2017-10-10 12:06
I would rather learn how to fly and take care of my spark correctly than have to pay for extra for stuff it doesn’t need as you say it’s not designed to fly miles doesn’t have the battery for it either but some around think it’s an inspire2 with the new Zenmuse x7 but they will only pay €500 quid for it, as far as my experience with it it’s a great little machine takes great photos easy launched easy landed plenty of safety features .
The first thing for a safe aircraft is a safe pilot .

Couldn't agree more.
X7? - is that what DJI are bringing out tomorrow do you think? I had a feeling its a new camera for the I2. I'm so tempted to get that quad but where to fly it!

Kloo Gee Posted at 2017-10-10 11:57
I think we all understand and can agree that we are talking about DJI's entry level drone and that for the price point, it is a technological marvel!  I think we can also all agree that the given hardware specifications of the Spark are likely largely in consideration of that price point.

I don't think anybody here is making any suggestions to change that hardware because those limitations are understood.  I think from my perspective, we are just suggesting DJI to utilize the given hardware they do have implemented to its fullest extent.

I’m sure none of us are here to save dji a sh#t load of money, the reality that I see is there is probably 500/600 people around this forum (spark) in the last week I have seen 1 maybe 2 cases of compass problems where gps was lost, but pilot recovered aircraft so I don’t think there are engineers working through the night on broken sparks.

Who is going to say no to a safety option that is not going to help avoid loosing his drone, and that includes dji I don’t believe for a minute that someone at dji hasn’t suggested a similar plan as it’s been discussed on these forums for years now, but while I don’t know the answer why they don’t or haven’t implemented I’m sure there will be good reason behind it , whether it doesn’t work whether they think it is not a safer option or physics says it just won’t work.

Since P2 I have seen amazing technology and safety improvements and dji have never been backward in coming forwards, and I’m almost certain that if their can be safer technology and it works out giving their customers better products and in turn making bigger profits for them we will see it as soon as it’s working and available.
So why have we not got it now my thinking is it’s not better than what is there at the moment, but when it is we will have it. And remember we haven’t seen any of these theory’s working yet is it for the same reason.

djiuser_Kf4iPA3 Posted at 2017-10-10 12:20
Like I said, take that argument up with DJI, because they're the ones saying you can stop any RTH procedure at will. Not me.

Here's what they say:

Critical battery is not a RTH mode - it is an emergency action whereby the drone will begin descending - it can be controlled by adding throttle on the left stick (in mode 2) to stop the descent as you manually fly the drone using the pitch, roll and yaw - it is all controlled manually though - nothing autonomous except the drone trying to land itself to prevent it hitting 0% power and crash landing. It has nothing to do with flying itself back to the home point (RTH).

ACW Posted at 2017-10-10 12:34
Critical battery is not a RTH mode - it is an emergency action whereby the drone will begin descending - it can be controlled by adding throttle on the left stick (in mode 2) to stop the descent as you manually fly the drone using the pitch, roll and yaw - it is all controlled manually though - nothing autonomous except the drone trying to land itself to prevent it hitting 0% power and crash landing. It has nothing to do with flying itself back to the home point (RTH).

You're right, good point. It lands right there, so it can't be considered an RTH procedure.
In that case, I'll change my previous statement - according to DJI, All RTH procedures without exception can be exited at will...

djiuser_Kf4iPA3 Posted at 2017-10-10 12:41
You're right. It lands right there, so it can't be considered an RTH procedure.
In that case, I'll change my previous statement - according to DJI, All RTH procedures without exception can be exited at will...

No, it doesn't land right there - it begins landing as I said and the descent is controlled manually using the throttle whist you manually fly it back to your position before it crashes.
And I wouldn't change your statement as you've already stated that failsafe RTH can NOT be exited before signal is reestablished - that being the exception.

ACW Posted at 2017-10-10 12:46
No, it doesn't land right there - it begins landing as I said and the descent is controlled manually using the throttle whist you manually fly it back to your position before it crashes.
And i wouldn't change your statement as you've already stated that failsafe RTH can NOT be existed before signal is reestablished - that being the exception.

True, but that's taking it out of context.
The point raised before was that in RTH, the user loses control of the AC. But if the user can stop any RTH at will, and resume manual control, it's not really loss of control, is it?
That's what I suggest in my proposed procedure too.
The case of lost connection doesn't have anything to do with the context of the argument. If there is no connection, ATTI wouldn't work either. If control isn't resumed, your drone is likely gone.

(in parentheses, I would say that if the drone has a procedure that moves it closer to you automatically, with or without compass, you have a higher chance to resume connection. Just saying...)

djiuser_Kf4iPA3 Posted at 2017-10-10 12:20
Like I said, take that argument up with DJI, because they're the ones saying you can stop any RTH procedure at will. Not me.

Here's what they say:

Think you will find that you manual tells only use RTH in emergency situation. It’s not the toy you think it is.

hallmark007 Posted at 2017-10-10 12:55
Think you will find that you manual tells only use RTH in emergency situation. It’s not the toy you think it is.

Actually I was looking for a quote like this in the last version of the manual (v1.4) and couldn't find one. If you can, feel free to quote it here, and I will stand corrected.
Smart RTH is not an emergency procedure, for one.
And it doesn't change my point - all RTH modes can be exited at any time at the user's choice, and manual control can be resumed. That's what I'm proposing as well.

djiuser_Kf4iPA3 Posted at 2017-10-10 12:52
True, but that's taking it out of context.
The point raised before was that in RTH, the user loses control of the AC. But if the user can stop any RTH at will, and resume manual control, it's not really loss of control, is it?
That's what I suggest in my proposed procedure too.

Sorry, but you made the statement regarding failsafe RTH so why add it if not relevant? I'm merely citing your own words...
Whilst the quad is engaged in RTH the flight path is fully autonomous - deactivating RTH no longer has the drone in RTH but before that deactivation the drone is not in your full control. Failsafe is the most common reason RTH activates and with no signal you have absolutely no control.
You may wish to look up flight autonomy which DJI introduced with the P4P. It sees the environment through the main sensor and records land marks in 3D imaging during it's route and returns home via that route by steering in line with the landmarks - thus bringing it closer to you on the previous flight path to regain signal quicker without a direct line RTH - this tech is already in my P4P and works brilliantly - you will never see it in the Spark though as too expensive.

djiuser_Kf4iPA3 Posted at 2017-10-10 12:59
Actually I was looking for a quote like this in the last version of the manual (v1.4) and couldn't find one. If you can, feel free to quote it here, and I will stand corrected.
Smart RTH is not an emergency procedure, for one.
And it doesn't change my point - all RTH modes can be exited at any time at the user's choice, and manual control can be resumed. That's what I'm proposing as well.

For one who is so concerned with safety, it beholds you to read the safety guidelines manual, you will see it on page 13.

I talked about it because being able to exit RTH mode was relevant to some sub-argument within our big argument about the RTH-without-compass procedure suggestion.
A lot has been said in this discussion, if you want to read back and understand the context of that point feel free. It may be a tedious undertaking... You have been warned...
And in the context of that, the possibility of having no control connection whatsoever on the aircraft is not relevant, because ATTI mode is not an option anyway.

Without reading back I get the impression your idea is for the drone to engage RTH and fly back closer to you when the compass/IMU/GPS is lost and then you turn off RTH manually when regained? Is that right? And to have this feature added to the Spark? Good luck if so.

ACW Posted at 2017-10-10 12:46
No, it doesn't land right there - it begins landing as I said and the descent is controlled manually using the throttle whist you manually fly it back to your position before it crashes.
And I wouldn't change your statement as you've already stated that failsafe RTH can NOT be exited before signal is reestablished - that being the exception.

I thought it might drop a little in price, but doesn’t look like it I’m very tempted just spend my trying to justify it,

I think we may see Zenmuse 7 tomorrow

ACW Posted at 2017-10-10 11:48
There will always be issues with any device or drone you purchase - I've seen videos on YT of a $10,000 I2 fall from the sky because the new prop locks stuck mid flight - all tech will have issues. My point is that we need to accept that these issues happen and even more so when you are talking about entry level - you get what you pay for and need to adjust to the product in hand. I fly all of my drones very differently in line with their limitations. There are probably hundreds of Sparks with minor faults, handfuls with major ones and thousands without any issues. I took my Spark out yesterday and no matter what I did the gesture control to start video recording wouldn't work - oh well! I just switched to manual flight, pressed record, switched to gesture and guess what - it recorded the rest of my active track in gesture control. It was probably due to the degree I raised my arm or something stupid and will probably work next time but I accept this is not a perfect drone and I didn't put a thread on here "why doesn't my gesture record work anymore". If DJI can implement firmware to improve flight safety then great and more of it but until then we all need to appreciate the limitations, adjust to those limitations and stop chasing our tails on here and pretending our £500 drone is faultless. I've never had an IMU failure or lost GPS though with my Spark and I for one don't believe in luck...

There is always an opposite side to consider, i own in excess of 20 drones, all my cheap drones come with fpv wifi using a phone or tablet, some even come with tapfly and rth based on the RC position (none have GPS), all have an RC, so spark with an RC is a $700 drone. I love doing mods to the cheap drones (the 249 gram challenge and everybody needs a hobby so this is mine) so i have a modified 5XUW a $100 drone, $26 to extend its range to 1.3 kilometers, $46 for the lightest 1080p dash cam with internal battery (on sale), $33 for vibration dampeners and a mini servo for gimbal control. The image quality achieved rivals the spark and in low wind conditions having no GPS is not an issue, i would love to add GPS but the drone is at 237 grams.
Now i have never ever had compass errors or interference with these cheap drones, you just fly them, no issues whatsoever. So lets tally the modded 5XUW, it comes to $205, so now what am I to expect from a $700 spark; as a minimum for the extra $495, a worry free flight experience but its just not able to provide that. I have modded the spark at 247 grams of course and fly it mostly in ATTI  and my photos will still be awesome. Is it worth the extra $495, yes if it works like promised, had the compass error to a point were it could not fly the spark at all no matter the location, but since i had problems with DJI assistant 2  recognizing the spark (another problem DJI did not solve for over a year) those records are lost and is a lost opportunity for DJI as a 100% compass failure rate is the ideal to solve the issue.