Flight report - https://app.airdata.com/flight/0a204bc8915d1ec4dffe476e360c8187/GENERALOverview
64 total flights with a total distance of 267,523 feet on my mavic pro 2.

My last flight ended with a lost mavic pro 2 in the water for reasons I cannot understand which leads me to believe it is a bug with the mavic pro 2?

I had a min and 26 secs of airtime, then my phone accidentally disconnected from the controller, and it stays in the air like it's done many other times when it disconnects from my phone. But this time, it just dropped like a rock shortly after my phone disconnected.

I love my mavic pro 2 and hope to get this replaced, but am horrified by the thought of this ever happening again and hope friends out there can help understand what happened on this flight.

Thank you for any help or advice you can provide.
Ryan

Sorry to hear about your issue, hope your flight record will procure some leads, strange that you only had 594 feet and 2 mins of flight time flown before the Mavic went down?...Good luck with your case with DJI, I hope you can be up and flying back in the air soon...

May I ask you what phone was used? Sorry for your case.

According to the log the drone didn't drop but landed.
What was your setting in case of signal loss? Landing? That is fatal over water.

Hello and good day Ryan. I am sorry to know what happened to your friends DJI Mavic 2 Pro and thank you for reaching out. Since this incident happened on the said drone I would recommend you contact our DJI support team to start up a ticket for the said drone at https://www.dji.com/support?site=brandsite&from=nav . We have the professional team who would do their best to fix and find out the reason of the incident and then the corresponding resolution would be provided. Thank for your support.

Hi Ryan,

Had a look at your flightlog. sorry for your loss.

2 times in the log you forced a landing to be executed.

Last time at 1m20s you pushed dow, craft at 4.7 meters height but VPS indicates 0.2 meter.(various between .2 and .7)In that situation software thinks that you want to land the craft!

Looks to me if you flew your mavic into the water.
Guess you only lost the video, as data is written into the file on your phone.

First ForceLanding at height 6.5 meters but VPS height indicates 0.3 or 0.4. (sensors dirty or wet or mis-functioning), but same as second one, conditions like you would normally land your craft. Hover above ground at low height and push stick down, once its landing you can release stick ; craft continue to land. (only stopped by 100% up)

Just after takeoff VPSheight was correct ; at 5.6s BARO height 6.5 and VPS height 8.0   (did you hand take off ?)After that VPS indicates nothing, at 9.2s BARO height 6.4 but VPSh 0.4.  So what happend in between? water drops to the sensor??

I think DJI should alter the software , at low height and with some forward speed and 100% stick down do ask the OP if he/she really wants to land the craft!!

BTW your lost connection setting = return to home.
cheers
JJB

JJB* Posted at 7-21 11:33
Hi Ryan,

Had a look at your flightlog. sorry for your loss.

Awesome find! I wish people would stop blaming DJI for their own error.

Lucas775 Posted at 7-21 13:02
Awesome find! I wish people would stop blaming DJI for their own error.

I do not think this is a user error. If i fly my drone above water at a height of 6 meters and i push down i expect thayt my drone will fly lower, and ofcourse i have my eyes at the drone and not at te app at the same time.

But in this scenario drone doesn`t lower at first, it waits a little and than execute a landing. (same as i land my craft manually above ground).  User is not aware that the software initiates a forcelanding, so is not aware that he has to stop by 100% up-stick. Releasing the sticks and think that drone will stabilize at that height...nope, it is just landing !

So IMO DJI should do something about this, 1) audio warning to indicate landing mode!  2) if flying with forward speed do no enter automatically landing mode but ask the OP if this is really what he wants.

cheers
JJB

There is an audible warning - it says "landing" if you have the volume up.

VPS gets confused over water which can trigger this - the manual states (or did the last time i checked) to turn off landing protection during over water flights for this reason.  I had mine initiate auto land from 50m altitude over the ocean once, VPS thought it was at 1m and i was descending (so down stick for 3 seconds or so).  I heard it say landing and cancelled it immediately.

JJB* Posted at 7-22 04:02
I do not think this is a user error. If i fly my drone above water at a height of 6 meters and i push down i expect thayt my drone will fly lower, and ofcourse i have my eyes at the drone and not at te app at the same time.

But in this scenario drone doesn`t lower at first, it waits a little and than execute a landing. (same as i land my craft manually above ground).  User is not aware that the software initiates a forcelanding, so is not aware that he has to stop by 100% up-stick. Releasing the sticks and think that drone will stabilize at that height...nope, it is just landing !

I don't really understand your logic regarding this not being a user error, since the aircraft did exactly what it's supposed to do: after one second of full negative throttle while less than a meter above the ground, it began landing. As far as prohibiting the landing when there's any lateral movement, that could make landing more difficult or even impossible in windy conditions when there's no way to keep a drone completely still. Also, I believe there is already a threshold above which it won't land even with negative throttle for one second, though in this case the aircraft was moving pretty slowly. I suppose we could debate what that threshold should be and the argument made that it's too low, but I'm guessing DJI picked what they believe to be a happy medium between safety and convenience.

It's unfortunate that this happened, but it clearly was a pilot error and not DJI's fault.

Gimpy Posted at 7-22 14:12
I don't really understand your logic regarding this not being a user error, since the aircraft did exactly what it's supposed to do: after one second of full negative throttle while less than a meter above the ground, it began landing. As far as prohibiting the landing when there's any lateral movement, that could make landing more difficult or even impossible in windy conditions when there's no way to keep a drone completely still. Also, I believe there is already a threshold above which it won't land even with negative throttle for one second, though in this case the aircraft was moving pretty slowly. I suppose we could debate what that threshold should be and the argument made that it's too low, but I'm guessing DJI picked what they believe to be a happy medium between safety and convenience.

It's unfortunate that this happened, but it clearly was a pilot error and not DJI's fault.

Hi Gimpy,

Your are right in your way, craft did exactly what the user requested.

But ; In this situation the user did not want to land his craft, he wanted to fly lower. Ofcourse if he knew the manual by head and understand all implications he had not made this move...and yes, disable settings in the app before would help too.How many people do fully understand all the settings and how to use them in a good way ?  For sure not many as you start with this hobby as a newbie.

Why should the software thinks that a user wants to land when flying with forward speed?? Not just a little fwd but just flying fwd.GPS Hovering on the spot with lots of wind is easy, have to check if speed is indicating, afaik it shows ground speed so always almost zero.
Not sure if the treshold value is set right, see if i can test this out flying low with speed and trying to get into a forced landing.

At his first forced landing speed was increasing, at entering landing drone speed was 3.9 m/s = 14kmh and increasing. Not really low value.So imo sofware should recognise this and stop landing mode, or at least give some kind of a warning beep.

In a normal user initiated landing imo the drone fly`s above ground, hovering at the spot at low height. 100% down stick means wait a little and enter landing mode.
So IMO DJI should add this fwd speed into the conditions to be met to initiate a manual forced landing, if done by DJI than i am curious about the treshold value.

Software must be made in a way that this kind of 'errors' could not be made. That`s why in bomber aircraft the release bomb switch is not close to a frequent use light switch etc.
Ergonmics in design is so important. In the app the RTH and Landing soft button are close to each other, wrong design! Few people on here have initiates a landing while thinking they did activate the RTH.  So DJI bring that landing btn to the right and leave the RTH btn on the left.


cheers
JJB

gnirtS Posted at 7-22 04:56
There is an audible warning - it says "landing" if you have the volume up.

VPS gets confused over water which can trigger this - the manual states (or did the last time i checked) to turn off landing protection during over water flights for this reason.  I had mine initiate auto land from 50m altitude over the ocean once, VPS thought it was at 1m and i was descending (so down stick for 3 seconds or so).  I heard it say landing and cancelled it immediately.

Hi gnirtS, guess i have to volume up a bit   or a hearing aid   

Mayby a double  beep-beep / beep beep  would do better to indicate that landing has started.

cheers
JJB

JJB* Posted at 7-22 23:39
Hi Gimpy,

Your are right in your way, craft did exactly what the user requested.

If it was going to hit someone or something, should it give warning first ? I’ve seen similar with newbies failing to let go when it is the best option.

hallmark007 Posted at 7-23 01:08
If it was going to hit someone or something, should it give warning first ? I’ve seen similar with newbies failing to let go when it is the best option.

oops, i do not really understand a) what your message has to do with my post to Gimpy and b) what hitting someone has to do with lower a craft low above the water.

But thinking about it ; mayby DJI can set an electric shock to the RC to alarm the OP that his/her drone is almost hitting something. You bring me some good ideas David, thanks.  

cheers
JJB

JJB* Posted at 7-23 02:04
oops, i do not really understand a) what your message has to do with my post to Gimpy and b) what hitting someone has to do with lower a craft low above the water.

But thinking about it ; mayby DJI can set an electric shock to the RC to alarm the OP that his/her drone is almost hitting something. You bring me some good ideas David, thanks.  

If I was trying to avoid hitting somebody or something by landing while craft is going forward, You say we should first get a warning first in the case we are not trying to land, I say that both safety of general public and property should always supersede the need to save craft.

In short I have the ability to immediately land my craft without having to go through any warning which might delay getting craft on the ground and in turn save someone from getting hurt or property being damaged.

You seem to be saying that it should be preferable to save craft first before endangering people or property, I just take the other view, considering there is already an existing warning there.

There is plenty of help in the manual as well as some very good advice in forums as well as in your safety  manual regarding flying over water.
You can’t always be holding drone users hands, it’s no harm for us to accept responsibility and learn from it and move on , it’s just my opinion.

hallmark007 Posted at 7-23 02:59
If I was trying to avoid hitting somebody or something by landing while craft is going forward, You say we should first get a warning first in the case we are not trying to land, I say that both safety of general public and property should always supersede the need to save craft.

In short I have the ability to immediately land my craft without having to go through any warning which might delay getting craft on the ground and in turn save someone from getting hurt or property being damaged.

Uhm, somehow i think my posts are  like a red rag for you as in a bull fight.

I never said or seem to be saying that craft safety becomes first over people safety!
Don`t know why you think that of me, must be that red rag; i have to react on this guy!

In a normal situation where users land there craft manually, its visible and no people in the way.

In situations that a user fly at low height forward and 100% down to lower, ofcourse there is nothing in the way (object or people) i think in 100% of the cases, as the user fly LOS and is seeing what he/she is doing.

Back to my statement ; if software recognises 100% down at low altitude with forward speed (and mayby increasing speed) ; don`t go into landing mode or stop the landing mode.

So IMO nothing to do with drone safey over people safety.

And yes fully agree with you about not holding drone user hands in every situation. But seen few times users get into forcelanding mode without wanted this and without knowing this. So a simple addition to the sw like a beep beep entering landing mode does the trick! It is normal for software to taken into account 'silly' actions from people using machinery, cars, etc, thus preventing that things goes utterly wrong.
Questions is only how deep software engineers wants to go, that`s why in developping sw user tests are so important. Writing behind a desk covers a lot but not all situations.

cheers
JJB

JJB* Posted at 7-22 23:39
Hi Gimpy,

Your are right in your way, craft did exactly what the user requested.

There's a simple solution for anyone who finds this behavior problematic: just turn off landing protection, and the aircraft won't auto-land after a second of full negative throttle while near the ground. In the meantime, the aircraft did what it was supposed to do.

JJB* Posted at 7-23 04:31
Uhm, somehow i think my posts are  like a red rag for you as in a bull fight.

I never said or seem to be saying that craft safety becomes first over people safety!

It seems if someone has a different opinion than you that somehow they are attacking your opinion, I simply made it known my opinion as to why we should not have any further warning to a command that is to land by user , a command like land should not be delayed anymore than is needed.
There are other failsafe on all dji craft including turning off landing protection and a beep to let you know it’s landing. How many more is required ?
I think the whole point on this thread is in simple terms, if you are going to take risks on low flying over water, inform yourself of the dangers what could go wrong and how to try avoiding this, if dji was to offer new craft for all risky flights, then who needs insurance who needs to read the manual , who needs to be responsible.

Your analogy regarding machines cars etc, well if I go full throttle with my car and break late, it’s highly likely I will crash , if I indicate right and I turn left I’m likely to crash but I have no safety for such a simple wrong movement, so you see there is nobody babying anyone with machines cars drones etc but the makers manufacturers expect those who use them to do so correctly and be responsible for their actions.

Lastly there are other people here who don’t necessarily agree with all you say and they rightly point this out and that’s the way forums of debate should be, they’re not personal .

Don't see why people are arguing here, take it to PMs, no need for it to be here. This case looks like user error, initiating a landing into water, flying low over water in the first place is a bad idea. If your unsure of the drones features i recommend you read over the manual as you may want to disable some features in certain environments and to meet certain needs, i believe it also warns users of flying above water. I believe the sensors can have errors due to the reflective and moving surface, but the main reason is probably due to the fact that if you crash in the water flight records on the drone can be lost and in some cases DJI would require this to prove fault or error and if these flight records are missing then DJI will list the case as no warranty due to the crash and lack of flight data.
I believe auto landing allows the use of Aileron and Elevator so the user can adjust during an auto landing.

I don't remember but I think I deactivated once the Landing Protection and then the drone started landing abruptly.

If that's the case and theoretically speaking,  then Landing protection should never be deactivated because it provides protection. DJI could consider assigning the auto landing function to a stick combination, therefore a landing would never happen unless the user really asked for it.   

jpap Posted at 7-23 07:18
I don't remember but I think I deactivated once the Landing Protection and then the drone started landing abruptly.

If that's the case and theoretically speaking,  then Landing protection should never be deactivated because it provides protection. DJI could consider assigning the auto landing function to a stick combination, therefore a landing would never happen unless the user really asked for it.

You're either remembering wrong or your drone had a problem that isn't widespread: turning off landing protection does not cause spontaneous, random landing to occur.

hallmark007 Posted at 7-23 05:47
It seems if someone has a different opinion than you that somehow they are attacking your opinion, I simply made it known my opinion as to why we should not have any further warning to a command that is to land by user , a command like land should not be delayed anymore than is needed.
There are other failsafe on all dji craft including turning off landing protection and a beep to let you know it’s landing. How many more is required ?
I think the whole point on this thread is in simple terms, if you are going to take risks on low flying over water, inform yourself of the dangers what could go wrong and how to try avoiding this, if dji was to offer new craft for all risky flights, then who needs insurance who needs to read the manual , who needs to be responsible.

Yep, we all may have different opinions and discuss that in a fair way. I like good discussions that way.And we all benefit from reading discussions and get more knowlegde about drone flying.

But think again when you or i started new to this hobby. How did we get all the knowledge to fly in all situations safe?  Reading the manual many times, watch YouTube videos and read on fora.

Afaik nothing is said in the manual about disable landing protection wehn planning to fly above water.
And why should i do this or why should i know this (as a newbie) ?  its called landing protection because its function is to protect landing....i just like to fly above water (as many do) and have no intention to land there :-)

The OP wanted to fly above water, and did not flew that low the first time, 6.5 meters is not taking too much risk in my mind.  BAD luck for the OP that VPS height was not indicating the correct value. Yes, if he had looked in the app he could have spotted that but i guess his eyes were at the drone.
Flying forward, increasing speed and lower the craft using 100% stick made the drone enter an autolanding mode. Not beacuse he wanted to land ofcourse, because of the software.Why 100% down stick, beacuse at first watching the drone it does not go down so more stick to see if it reacts on what i wants.

All the cautions in the manual about the performance fdw and down vision systems just say the the ac should be operate with great caution.
Nothing specific what could happen in any situiation, ofcourse not so how to know the story behind this.

We all may have our own opinion about this flight, i believe my point is a valid one.  Hope that DJI will jump in and change some pre-conditions for entering this landing mode.

Happy many safe landings to you all, no hard feeling this side. Its just my opinion.

cheers
JJB

JJB* Posted at 7-23 08:38
Yep, we all may have different opinions and discuss that in a fair way. I like good discussions that way.And we all benefit from reading discussions and get more knowlegde about drone flying.

But think again when you or i started new to this hobby. How did we get all the knowledge to fly in all situations safe?  Reading the manual many times, watch YouTube videos and read on fora.

You need to look at the log again. When the last autolanding occurred, he wasn't 6.5 meters above the water but just one foot (well under one meter). If you're flying just above the surface and pull all the way back on the throttle for a full second, you should expect your drone to get wet one way or another. If he hadn't flow enough over the ground to realize how landing protection's autolanding works, then his biggest mistake was flying over the water (and that close to the surface of it) to begin with without having more experience first and ignoring advice in the manual in the process. Regardless of how you try to spin it, this was a case of pilot error. You are, of course, entitled to your opinion, but I think it's extremely unlikely DJI will consider this a defect / warranty case, and if they did they'd need to apologize to many other pilots who didn't get the same treatment in the past.

JJB* Posted at 7-23 08:38
Yep, we all may have different opinions and discuss that in a fair way. I like good discussions that way.And we all benefit from reading discussions and get more knowlegde about drone flying.

But think again when you or i started new to this hobby. How did we get all the knowledge to fly in all situations safe?  Reading the manual many times, watch YouTube videos and read on fora.

I’m afraid your assuming that most of us just took off without any knowledge or prep or reading manuals or taking advice, I see it regularly here people asking for best advice when flying over water, it’s called being responsible, if we tell them don’t bother just go fly if it goes wrong you’ll get a new craft, it’s clear that there are enough warnings failsafe if dji were to introduce a completely safe drone with no challenge that could fly by itself, then it’s likely that no one would be interested because as a hobby it would have no challenge .
When someone crashes because of lack of knowledge or ignorance rewarding him with a new drone will not teach him anything.
The truth is the tools are there to get from A to B we must learn them to achieve this and if we have to learn the hard way then it will make us stronger , and yes if dji can improve as they clearly have then great .

Gimpy Posted at 7-23 08:48
You need to look at the log again. When the last autolanding occurred, he wasn't 6.5 meters above the water but just one foot (well under one meter). If you're flying just above the surface and pull all the way back on the throttle for a full second, you should expect your drone to get wet one way or another. If he hadn't flow enough over the ground to realize how landing protection's autolanding works, then his biggest mistake was flying over the water (and that close to the surface of it) to begin with without having more experience first and ignoring advice in the manual in the process. Regardless of how you try to spin it, this was a case of pilot error. You are, of course, entitled to your opinion, but I think it's extremely unlikely DJI will consider this a defect / warranty case, and if they did they'd need to apologize to many other pilots who didn't get the same treatment in the past.

Nah, second forcelanding started at baro height 4.9 meters, VPS indicating 0.5.
VPS not correct value.
Noboby would apply 100% down stick flying at 0.5 meter while having the drone visible.

I never spoke about a DJI warranty claim.

Not trying to spin but seems hard for others to discuss with arguments.
How the heck does a newbie knows that autolanding will be active with a wrongly indicated VPS height and flying with fwd speed?

To be honest, i did not know until i saw some logs on this forum showing this happening.
Guess i am not the only one, that`s why this is a great forum for all readers!

cheers
JJB

hallmark007 Posted at 7-23 08:57
I’m afraid your assuming that most of us just took off without any knowledge or prep or reading manuals or taking advice, I see it regularly here people asking for best advice when flying over water, it’s called being responsible, if we tell them don’t bother just go fly if it goes wrong you’ll get a new craft, it’s clear that there are enough warnings failsafe if dji were to introduce a completely safe drone with no challenge that could fly by itself, then it’s likely that no one would be interested because as a hobby it would have no challenge .
When someone crashes because of lack of knowledge or ignorance rewarding him with a new drone will not teach him anything.
The truth is the tools are there to get from A to B we must learn them to achieve this and if we have to learn the hard way then it will make us stronger , and yes if dji can improve as they clearly have then great .

I see 2 kind of people on this forum, some as newbies; one group is asking good questions and read manuals etc. As you said ; responsible people and preparing themselves for safe flights.
Second group thinks that flying drones is playing with fun toys, do have little knowledge and make silly mistakes, often ends in loosing a craft.
I have no % for both groups.
So not sure why you think i assume always in the negative....On this forum we discuss and tell often to do after take off checks, in how many logs do we see this back ??  They are so important for a good flight!



cheers
JJB

JJB* Posted at 7-23 11:51
I see 2 kind of people on this forum as newbies, one group is asking good questions and read manuals etc. As you said ; preparing for safe flights.

Second group thinks that flying drones is playing with fun toys, do have little knowledge and make silly mistakes, often ends in loosing a craft.

Those who read the manuals ask the questions plan their flights should get the Apple from the teacher not those who couldn’t be arsed lol....

hallmark007 Posted at 7-23 12:09
Those who read the manuals ask the questions plan their flights should get the Apple from the teacher not those who couldn’t be arsed lol....

yep, i spent years throwing apples at my students!  and they were happy about that (well, bit later in times ofcourse, as they and mayby we are great adult-children)

JJB* Posted at 7-23 11:51
I see 2 kind of people on this forum, some as newbies; one group is asking good questions and read manuals etc. As you said ; responsible people and preparing themselves for safe flights.
Second group thinks that flying drones is playing with fun toys, do have little knowledge and make silly mistakes, often ends in loosing a craft.
I have no % for both groups.

"Nah, second forcelanding started at baro height 4.9 meters, VPS indicating 0.5."

At 1:20.3 -- when the operator pulls back on the throttle -- the VPS height is 1 foot (0.3 meters).

"VPS not correct value. Noboby would apply 100% down stick flying at 0.5 meter while having the drone visible."

In other words, your opinion that this wasn't pilot error is based on your own SPECULATION about how a person would behave, as opposed to what the log actually shows. I see.

In any case, something you seem to have overlooked -- or just didn't know -- is that it's the VPS height that determines whether auto-landing is initiated, so if VPS believes that it's 1 foot from the surface, it's going to start landing regardless of whether or not it actually is that close. DJI is very well aware of the fact that VPS isn't always 100% accurate, mentions that it can be inaccurate in the manual, and warns about flying over water.

Gimpy Posted at 7-23 12:33
"Nah, second forcelanding started at baro height 4.9 meters, VPS indicating 0.5."

At 1:20.3 -- when the operator pulls back on the throttle -- the VPS height is 1 foot (0.3 meters).

mayby a graphic will make it clear.

If you read al my posts you will read that VPS height started to indicate in the correct way at start of the flight. later in flight indication went faulty.
See the first baro blue line and vps line are in sync, later just rubbish.

So height was not 0.5 meters actual but more. its all in the log.

And no, your end conclusion is wrong. Not based on speculation but just interpreting data.
I think sometimes exprerienced flyers judge to much from their own level and knowledge and not from a starter level in drone flying.

cheers
JJB

JJB* Posted at 7-23 12:43
mayby a graphic will make it clear.

If you read al my posts you will read that VPS height started to indicate in the correct way at start of the flight. later in flight indication went faulty.

Your own screenshot shows exactly what I said: that when he applied full negative throttle for the last time, the VPS reading showed him just a foot above the surface. VPS is used to to determine whether to start auto-landing, which it did in this case. As far as what the height actually was, you don't know and neither do I: we only can see what's in the log, which indicates that the relevant (VPS) height was 1 foot. The barometric altimeter height is irrelevant because it isn't and can't be used as a basis for determining whether to auto-land due to its low accuracy.

What we do know is that he triggered an auto landing by pulling back on the throttle for a full second. We also know that he didn't reverse / abort that landing by pushing forward on the throttle, so even if your speculation about his height being greater than what VPS showed is correct, that means he had even more time to abort the landing but still didn't do it. Either way it winds up being pilot error, at least according to any reasonable definition of the phrase. If you prefer to have and use your own private and very narrow definition of "pilot error", that's your choice, but ultimately the only opinion that really matters is DJI's.

JJB* Posted at 7-23 11:46
Nah, second forcelanding started at baro height 4.9 meters, VPS indicating 0.5.
VPS not correct value.
Noboby would apply 100% down stick flying at 0.5 meter while having the drone visible.

I suspect DJIs response to that would be the manual states vision system maybe unreliable and use caution when flying over water.  (Page 22 roughly).

Its backed off from the Mavic 1 manual where it stated to turn off downward sensors over water but it still says unreliable and use caution over water.  So i suspect they'll just point to that and not replace.

Gimpy Posted at 7-23 12:49
The screenshot from your own utility shows exactly what I said: that when he applied full negative throttle for the last time, the VPS reading showed him just a foot above the surface. VPS is used to to determine whether to start auto-landing, which it did in this case. As far as what the height actually was, you don't know and neither do I: we only can see what's in the log, which indicates that the relevant (VPS) height was 1 foot. The barometric altimeter height is irrelevant because it isn't and can't be used as a basis for determining whether to auto-land due to its low accuracy.

At this point, you're refusing to acknowledge the facts you're presenting yourself (i.e., the VPS height in the above graph), so I can only guess that you're embarrassed about being wrong and trying to save face. I suspect that stubborn refusal to acknowledge what's apparent to everyone else may have be part of the reason for the negative reactions to your posts you mentioned earlier in this thread -- but like your guess about what the Mavic's real altitude was here, that's just speculation on my part.

One thing we nearly always see is vps low to the ground is almost always correct and opposite can be said for barometer.

Hi,

Thank you everyone for the thoughts. Before I go into it, just wanted to share i am hard of hearing, about 75% hearing loss and I use a hearing aid, so i didn't realize there was an audio cue till I read the comments in this thread.

As for the data showing that I triggered the landing, it does make sense but one thing the data shows is my phone disconnected from the controller due to something wrong with the insert plug section, has to be at an exactly right position to transfer data for some reason, I'm getting a new phone which should replace that problem. After it disconnected, I looked at the drone and saw it was hovering in a good spot, so i decided to try to disconnect and reconnect my phone and get the app up and running again, this took me I think 20 ish seconds and think during this time, the drone did an auto-landing, before my phone reconnected, and was confused but now after seeing the data. is it possible that I accidentally hit the controller joysticks itself to initiate the landing? i admit right out of the box i used my phone as the main input all the way, and never mastered the controller itself which may have backfired on me now that i think about it.

Again, does seem like my fault and have no choice but to just accept my loss and buy another one. and the new phone I plan to buy should help with avoiding that type of issue again, but realize why can't DJI somehow recognize a disconnected phone at a weird time may mean issues, and locks the controller to an extent for a few secs? just spitballing ideas here, I will definitely be more careful in the future.

Thanks again everyone for helping me understand what happened to my drone, may it rest in peace.

Ryan H-c Posted at 7-23 20:18
Hi,

Thank you everyone for the thoughts. Before I go into it, just wanted to share i am hard of hearing, about 75% hearing loss and I use a hearing aid, so i didn't realize there was an audio cue till I read the comments in this thread.

Controller is working when phone is disconnected, it cannot lock when phone is disconnected as main reason is that you can still control your drone even without phone when you see it.

I strongly suggest you to buy version with smartcontroller and you will forget issues with disconnected phone forever. Firmware is little bit buggy so far, some minor things but from stability and signal point of view it is great. I think that firmware will be tunedup in near future to address all concerns users have so far.

Gimpy Posted at 7-23 12:49
Your own screenshot shows exactly what I said: that when he applied full negative throttle for the last time, the VPS reading showed him just a foot above the surface. VPS is used to to determine whether to start auto-landing, which it did in this case. As far as what the height actually was, you don't know and neither do I: we only can see what's in the log, which indicates that the relevant (VPS) height was 1 foot. The barometric altimeter height is irrelevant because it isn't and can't be used as a basis for determining whether to auto-land due to its low accuracy.

What we do know is that he triggered an auto landing by pulling back on the throttle for a full second. We also know that he didn't reverse / abort that landing by pushing forward on the throttle, so even if your speculation about his height being greater than what VPS showed is correct, that means he had even more time to abort the landing but still didn't do it. Either way it winds up being pilot error, at least according to any reasonable definition of the phrase. If you prefer to have and use your own private and very narrow definition of "pilot error", that's your choice, but ultimately the only opinion that really matters is DJI's.

Gimpy, never mind my arguments, you will not try to understand.

Agree on : VPS height is used for the autolanding, not the baro height.
VPS sensor information above reflective surfaces water is not always correct, like in this log, mayby the sensor got wet and that`s why we see it indicating with erratic values.

If drone was actually flying at 1 feet, why did it take 8.3 seconds to reach the water??  Just use landing speed in m/s and find the estimated begin height  ;-) Not 1 feet i can tell you.
But i leave it here, proof is all in the log but you have to see it.

Baro height may be not that accurate as Hallmark says, on the contrary i must say. Mayby the acutal value not but the data in sequence is accurate. And as the largest baro changes happens at low altitude the sensor can detect that better than at high altitude, imo the sensor is not that bad. Look at all the flightlogs i have seen, for sure my MA baro height is accurate.

Bad luck for the Op that his phone connection went bad, still in control so 100% up had killed the autolanding or partly up had stop the down movement.Pilot error? well, if OP isn`t yet that experienced than it was not a good idea to fly above the water. But imo software did not help him either.
So 50 / 50 for the mishap. (OP/DJI)

cheers
JJB

hallmark007 Posted at 7-23 15:33
One thing we nearly always see is vps low to the ground is almost always correct and opposite can be said for barometer.

Try it over water.  The VPS is incredibly inaccurate over water.  If it displays at all.
Theres a reason DJI say dont trust it near the surface.

hallmark007 Posted at 7-23 15:33
One thing we nearly always see is vps low to the ground is almost always correct and opposite can be said for barometer.

True indeed, low flying over ground with enought reflective material = good VPS height signal
Not always True flying above water.
Baro height not that bad, sometimes it varies 10 feet from start data to end of the flight, this often when craft went up to some height (> 300 feet)
But when in a flight the height variations are less, than less deviation begin - end flight and data in sequence trustworthy.

cheers
JJB

gnirtS Posted at 7-24 01:03
Try it over water.  The VPS is incredibly inaccurate over water.  If it displays at all.
Theres a reason DJI say dont trust it near the surface.

As someone who regularly flies over water and within altitude of 30 ft, VPs is almost more accurate than barometer, yes if you flying over fast moving water or ocean waves under 15 feet there can be incorrect measurements but barometer will not be more accurate . Yes when flying below 15 ft best to turn off VPS but also to fly VLOS .

JJB* Posted at 7-24 01:00
Gimpy, never mind my arguments, you will not try to understand.

Agree on : VPS height is used for the autolanding, not the baro height.

You're the one who's ignored what everyone else is trying to tell you about flying over water, about VPS altitude, and the barometer altitude. You also keep making up things you don't really know but insist you do, such as your new (and incorrect) claim that the barometer can only be off by 10 feet. You really need to learn to distinguish your speculation and opinions from fact; they aren't always the same. Just like I suspect that the reason you're still arguing about this is because you're embarrassed about being wrong regarding pilot error and want to save face; that's an example of something that's an opinion but may or may not be a fact.

Anyway, the OP has accepted responsibility for the crash, which is to his credit. Again, it's unfortunate that this happened, but telling him it wasn't his fault when it really was doesn't help anyone.