Map Pilot P4P Crash today

[SantoPilot]

I purchased a P4P Friday, replacing a Phantom 4 I'll use for backup (The P4 is in transit to DJI for camera issue - red/white pixels on vids/pics so grabbed a P4P while the P4 gets fixed). I got the new P4P home, did all the firmware upgrades, etc. Everything charged and ready for a Saturday test flight. Flew a couple of times on Saturday afternoon (morning storms) and had no issues, tested RTH, etc. Sunday I went to a job to survey a site and used Map Pilot. This flight would take 3 batteries, as usual, I have flown it a few times. I set everything up, P4P took off, did the first leg no problems (FLY008.DAT). Landed for the 2nd leg, changed battery, auto takeoff, P4P fell out of the sky (FLY009.DAT) minutes later. P4P was at 398 feet for this Mission. I was watching it when it just fell from the sky in to dense area of shrubs, stickers, vines, etc.. I tried RTH and such, nada. This happened very fast. After a few hours, found the P4P and started the research. I converted the DAT files to CSV, will open a support incident tomorrow with DJI. With all the data exported, I wanted a 2nd set of eyes to take a look as well. I am thinking a hardware failure on the P4P, curious what anyone else sees. I added the FLY008.DAT (first flight, no issues), then the FLY009.DAT (crash flight). I also added the Map Pilot csv files (I don't see much in there for help), the flight path and last few photos the P4P took. I've never had a Phantom just fall out of the sky like this. It's busted up now (gimble, case, gear, etc.). I am crossing my fingers it is hardware malfunction and a warranty item. Guess I'll find out more in a few days as I imagine I'll have to send the P4P and not just the logs to DJI. Note that since I used Map Pilot and not DJI Go app, there are not the txt files I read about, just the DAT's from the Drone itself.

Files can be found shared via OneDrive - P4P

Files are rather large and they might still be uploading.

Thanks to anyone that takes a stab at this!

 

[DocDiggler]

Oh no!
Not another one

 

[BudWalker]

I purchased a P4P Friday, replacing a Phantom 4 I'll use for backup (The P4 is in transit to DJI for camera issue - red/white pixels on vids/pics so grabbed a P4P while the P4 gets fixed). I got the new P4P home, did all the firmware upgrades, etc. Everything charged and ready for a Saturday test flight. Flew a couple of times on Saturday afternoon (morning storms) and had no issues, tested RTH, etc. Sunday I went to a job to survey a site and used Map Pilot. This flight would take 3 batteries, as usual, I have flown it a few times. I set everything up, P4P took off, did the first leg no problems (FLY008.DAT). Landed for the 2nd leg, changed battery, auto takeoff, P4P fell out of the sky (FLY00(.DAT) minutes later. P4P was at 398 feet for this Mission. I was watching it when it just fell from the sky in to dense area of shrubs, stickers, vines, etc.. I tried RTH and such, nada. THis happened very fast. After a few hours, found the P4P and started the research. I converted the DAT files to CSV, will open a support incident tomorrow with DJI. With all the data exported, I wanted a 2nd set of eyes to take a look as well. I am thinking a hardware failure on the P4P, curious what anyone else sees. I added the FLY008.DAT (first flight, no issues), then the FLY009.DAT (crash flight). I also added the Map Pilot csv files (I don't see much in there for help), the flight path and last few photos the P4P took. I've never had a Phantom just fall outmof the sky like this. It's busted up now (gimble, case, gear, etc.). I am crossing my fingers it is hardware malfunction and a warranty item. Guess I'll find out more in a few days as I imagine I'll have to send the P4P and not just the logs to DJI. Note that since I used Map Pilot and not DJI Go app, there are not the txt files I read about, just the DAT's from the Drone itself.

Files can be found shared via OneDrive - P4P

Files are rather large and they might still be uploading.

Thanks to anyone that takes a stab at this!

I looked at this very briefly earlier. I'm on my iPad and will do a more complete analysis later.

There was an abrupt power loss at the very end of the .dat. There a couple of theories about why this happens. 1 loose battery, 2 defective battery.

If someone else wants to look I suggest using datcon first. Set samplerate to max and set time interval to last few secs. Then look at motorVolts

 

[SantoPilot]

I noticed Compass reporting things (SPEED_ERROR_LARGE, GPS_POSITION_NONMATCH) and gpsHealth in CsvView jumping all over in the end and [L-RC]craft ctrl failed!!! type entries. When I found the P4P, battery was still in but dislodged from impact I assumed. When I got home to power it up to retrieve the logs, I used the same battery. I am no means an expert in these logs, thanks for the initial light analyisis.

Update: Opened a case with DJI today and sent them DAT files. P4P is now in transit to Cerritos, CA

 

[BudWalker]

I noticed Compass reporting things (SPEED_ERROR_LARGE, GPS_POSITION_NONMATCH) and gpsHealth in CsvView jumping all over in the end and [L-RC]craft ctrl failed!!! type entries. When I found the P4P, battery was still in but dislodged from impact I assumed. When I got home to power it up to retrieve the logs, I used the same battery. I am no means an expert in these logs, thanks for the initial light analyisis.

Update: Opened a case with DJI today and sent them DAT files. P4P is now in transit to Cerritos, CA

You're right, it's not a complete loss of power. I should know better by now. I was in a hurry and misread the data. Anyway, I took a closer look at FLY009 and it looks to me like there was a rightBack propulsion issue. At time 22 the P4P began to spin clockwise and tumble (starting with pitch up). Both of these are consistent with a rightBack propulsion issue. The nonGPSCause and gpsHealth errors were caused by the spinning and tumbling

In addition, the rightBack motor speed decreased and the rBack motor current increased

The increased current also caused a sag in the voltage, which in my haste, led me to believe there was a power interruption.

Does the right back spin as the others?

 

[SantoPilot]

You're right, it's not a complete loss of power. I should know better by now. I was in a hurry and misread the data. Anyway, I took a closer look at FLY009 and it looks to me like there was a rightBack propulsion issue. At time 22 the P4P began to spin clockwise and tumble (starting with pitch up). Both of these are consistent with a rightBack propulsion issue. The nonGPSCause and gpsHealth errors were caused by the spinning and tumbling
View attachment 84532
In addition, the rightBack motor speed decreased and the rBack motor current increased
View attachment 84529
View attachment 84530

The increased current also caused a sag in the voltage, which in my haste, led me to believe there was a power interruption.

Does the right back spin as the others?

The P4P is boxed and shipped to DJI, I didn't fire the motors back up to see if they would. Thanks for the review, I really appreciate it! I was thinking the rightBack look suspect to me in my limited knowledge of these logs. Makes sense at the tumble GPS and such would show strangeness. One thing I didn't mention, when the P4P was flying, on screen it was showing drastic speed changes. One second it's cruising around 19 mph, the next second 8.x mph, back and forth. I was thinking maybe it was winds aloft being that high and the P4P adjusting so I dismissed as a problem. That was around 20/30 seconds or so before it fell from the sky.

 

[BudWalker]

Yeah, I see that.

But, I suspect that these slow downs are part of the mission that it was runnning under. Perhaps you had it set to take a photo at those spots? It also slows down when the P4P changes direction.

Since you are already an accomplished CsvView bring up the GeoPlayer and a SigPlayer like above and you can see where it slows down. Just move the value tracker in the SigPlayer to those times where Vel:H slows down and you can see where the P4P was.

 

[SantoPilot]

Yeah, I see that.

But, I suspect that these slow downs are part of the mission that it was runnning under. Perhaps you had it set to take a photo at those spots? It also slows down when the P4P changes direction.
View attachment 84543
Since you are already an accomplished CsvView bring up the GeoPlayer and a SigPlayer like above and you can see where it slows down. Just move the value tracker in the SigPlayer to those times where Vel:H slows down and you can see where the P4P was.
View attachment 84544

Hey that's cool trick, thanks for the tip. I reviewed the FLY008.DAT as well to compare, shows the same slow downs. Map Pilot automatically sets where/when to take photos during a Mission.

 

[SantoPilot]

You're right, it's not a complete loss of power. I should know better by now. I was in a hurry and misread the data. Anyway, I took a closer look at FLY009 and it looks to me like there was a rightBack propulsion issue. At time 22 the P4P began to spin clockwise and tumble (starting with pitch up). Both of these are consistent with a rightBack propulsion issue. The nonGPSCause and gpsHealth errors were caused by the spinning and tumbling
View attachment 84532
In addition, the rightBack motor speed decreased and the rBack motor current increased
View attachment 84529
View attachment 84530

The increased current also caused a sag in the voltage, which in my haste, led me to believe there was a power interruption.

Does the right back spin as the others?

Just to give an update, DJI sent me a repair bill for this P4P of $194.00. I sent DJI info from what you saw asking for an explanation as to a Pilot error. I called DJI support (and email), turns out an analysis was not done on the logs. Since I requested this they will now look. When I was givien a case I also uploaded the DAT files per DJI Assistance app. I'll update when I know more. Seems a bit sneaky to me not to auto review logs unless a customer requests. I guess they save $$$ that way as some folks who do not know better just pay.

Thought maybe folks would be curious to the outcome, I'll keep this thread updated.

 

[SantoPilot]

Got a response from DJI (I had to push for an escalation). DJI performed a flight log analysis, Map Pilot logs and the DAT files on the P4P were reviewed. They concluded that the Map Pilot app provided a false reading to the sensors that triggered the P4P to think it was close to the ground. Thus the P4P engines began a shutdown as if it had landed.

I am paying for the repairs, in one way glad it was not a hardware failure on the P4P part. Since that flight, I found DJI recently released their own app DJ GS Pro for such type flights. I had a problem with Map Pilot once a while back on such a mission, but it was easily recoverable situation. Map Pilot is now deleted from my library of options.

 

[John Locke]

Got a response from DJI (I had to push for an escalation). DJI performed a flight log analysis, Map Pilot logs and the DAT files on the P4P were reviewed. They concluded that the Map Pilot app provided a false reading to the sensors that triggered the P4P to think it was close to the ground. Thus the P4P engines began a shutdown as if it had landed.

I am paying for the repairs, in one way glad it was not a hardware failure on the P4P part. Since that flight, I found DJI recently released their own app DJ GS Pro for such type flights. I had a problem with Map Pilot once a while back on such a mission, but it was easily recoverable situation. Map Pilot is now deleted from my library of options.

You should send DJI's conclusion to Map Pilot and ask them to reimburse you. Otherwise you'll post this everywhere (beyond PhantomPilots). It would be interesting to hear their response. I assumed DJI would blame the NON DJI software for the problem, whether it's true or not it causes the motor shutdown. I think DJI will blame anything that's not DJI for problems with their craft, sometimes legit, sometimes questionable. They told me that even a sticker wrap on the craft voids the warranty. But hey, I still love their drones.

 

[JWH]

You should send DJI's conclusion to Map Pilot and ask them to reimburse you. Otherwise you'll post this everywhere (beyond PhantomPilots). It would be interesting to hear their response. I assumed DJI would blame the NON DJI software for the problem, whether it's true or not it causes the motor shutdown. I think DJI will blame anything that's not DJI for problems with their craft, sometimes legit, sometimes questionable. They told me that even a sticker wrap on the craft voids the warranty. But hey, I still love their drones.

Thanks for the post about stickers voiding the warranty. I didn't know that. The only problem that I could see with stickers is if they had anything metallic in them.

I do enjoy the P4. So far I have only had one close call, which was entirely my fault. I am always reading these forums and learning from other's misfortune.

 

[SantoPilot]

You should send DJI's conclusion to Map Pilot and ask them to reimburse you. Otherwise you'll post this everywhere (beyond PhantomPilots). It would be interesting to hear their response. I assumed DJI would blame the NON DJI software for the problem, whether it's true or not it causes the motor shutdown. I think DJI will blame anything that's not DJI for problems with their craft, sometimes legit, sometimes questionable. They told me that even a sticker wrap on the craft voids the warranty. But hey, I still love their drones.

I sent info to Map Pilot today, also attached logs to the submission and asked them for reimbursement. This is what DJI sent me via email below. DJI did actually call me and left a voice mail stating the same. I was also told that a battery over 6 months old used in a problem scenario also voids warranty, but who knows from DJI support, agent might have just been talking about things he didn't know about -

Repair CAS-745091-Q2P0D3 Was determined Non-Warranty due to the information that was extracted from flight records.
Technician found pilot using a third party app which we were able to retrieve flight records from.
VPS sensors on the aircraft were affected by false readings provided by via the third party app.
VPS sensors upon receiving this signal caused the aircraft to detect close enough ground ending up in motor shut off while in flight.

 

[John Locke]

I sent info to Map Pilot today, also attached logs to the submission and asked them for reimbursement. This is what DJI sent me via email below. DJI did actually call me and left a voice mail stating the same. I was also told that a battery over 6 months old used in a problem scenario also voids warranty, but who knows from DJI support, agent might have just been talking about things he didn't know about -

Repair CAS-745091-Q2P0D3 Was determined Non-Warranty due to the information that was extracted from flight records.
Technician found pilot using a third party app which we were able to retrieve flight records from.
VPS sensors on the aircraft were affected by false readings provided by via the third party app.
VPS sensors upon receiving this signal caused the aircraft to detect close enough ground ending up in motor shut off while in flight.

Wow, that's a bad flaw in the MapPilot software, assuming what DJI is saying is true.. It will be interesting to see if MapPilot has ever heard of this happening elsewhere, and if they'll take responsibility. But the part about the battery age, yes DJI will take old batteries as an exception to warranty coverage if it's more than 6mos old, which is the warranty length of the battery.

 

[BudWalker]

..... They concluded that the Map Pilot app provided a false reading to the sensors that triggered the P4P to think it was close to the ground. Thus the P4P engines began a shutdown as if it had landed.
.........

This is total nonsense. The tablet app doesn't provide data for any of the sensors on the AC. Data originates from sensors. This logs "review" is just an obfuscation in the hope that the consumer gets weary and goes away.

The motors were not shutting down as if the P4P had landed. Only the rightBack motor was running slower and that went on for at least 6 seconds. Also, if the P4P thought it had landed, or was landing, that would show up in the eventLog stream. It does not show up in the eventLog stream.

 

[SantoPilot]

As expected, no response from Drones Made Easy with the MapPilot software. I'll hit them on social (facebook, twitter, etc.) to light it up a bit to see what happens. I do grow weary on this and am about done. I paid for the repairs, have some side jobs to get done. Not sure if it's worth the fight at this point. I was hoping the 2 companies would possibly talk, email, or however and not finger point. It's not the money on repairs that is the issue, it's the shoddy support or lack of accountability from companies that drives me. When or if I hear back, I'll update the thread.

 

[John Locke]

As expected, no response from Drones Made Easy with the MapPilot software. I'll hit them on social (facebook, twitter, etc.) to light it up a bit to see what happens. I do grow weary on this and am about done. I paid for the repairs, have some side jobs to get done. Not sure if it's worth the fight at this point. I was hoping the 2 companies would possibly talk, email, or however and not finger point. It's not the money on repairs that is the issue, it's the shoddy support or lack of accountability from companies that drives me. When or if I hear back, I'll update the thread.

The next letter should say something like, "I would appreciate a reply to the issue I have brought to your attention, it's the least you could do to clear up this matter". Then giving them the benefit of the doubt, "If you are still looking into this with DJI, at least let me know you're working on it". Then at the end say "Getting no response implies you know this is a defect in your software" and then let them know that "no response will lead to letting the public know of the flaw that DJI claims is in the MapPilot software on social media"

 

[SantoPilot]

Got a respone (Social media is at least good for a quick response) from DronesMade Easy. First thing they call out - "Everytime Map Pilot starts up it lets you know you are using at your own risk." I laughed at that statement, I know what my screens say when I use use apps.

Anyways, Drones Made Easy then said "For a catastrophic failure like this to occur it is likley that there was a hardware/sensor failure on the aircraft. There is literally no signal or message that the SDK can send that would make the aircraft drop from the sky." Yep, I get this as well, have written 50+ apps myself and understand the SDK's to know what DJI was doing what BudWalker stated earlier. Drones Made Easy did not review any logs BTW, not that it matters really.

I'll forward to DJI, but thats it, I'm done and am moving ahead. I appreciate all the help and comments, this is a time waster now for me.

 

[BudWalker]

Got a respone (Social media is at least good for a quick response) from DronesMade Easy. First thing they call out - "Everytime Map Pilot starts up it lets you know you are using at your own risk." I laughed at that statement, I know what my screens say when I use use apps.

Anyways, Drones Made Easy then said "For a catastrophic failure like this to occur it is likley that there was a hardware/sensor failure on the aircraft. There is literally no signal or message that the SDK can send that would make the aircraft drop from the sky." Yep, I get this as well, have written 50+ apps myself and understand the SDK's to know what DJI was doing what BudWalker stated earlier. Drones Made Easy did not review any logs BTW, not that it matters really.

I'll forward to DJI, but thats it, I'm done and am moving ahead. I appreciate all the help and comments, this is a time waster now for me.

I must be missing something. I have nothing to do with Drones Made Easy, never tried Map Pilot and don't know anything about it beyond what I've read here. It would seem that you've accepted that Map Pilot could not have caused this incident? What then do you expect the Map Pilot vendor should do?

 

[SantoPilot]

I must be missing something. I have nothing to do with Drones Made Easy, never tried Map Pilot and don't know anything about it beyond what I've read here. It would seem that you've accepted that Map Pilot could not have caused this incident? What then do you expect the Map Pilot vendor should do?

No I am not saying Map Pilot was the cause. Each points the finger at the other, nor that anyone is associated with either one. From what you said earlier, data originates from the sensors, I very much doubt the SDK lets you send data to those sensors. When DJI pointed at Map Pilot, I wanted their input (we all knew what they would kind of say and they confirmed) and sent the Map Pilot email to DJI. I don't expect anything from DJI as they have no way to analyze data at the front line and it appears they don't have or use escalation path. I asked that DJI to escalate to their SDK debugging team for confirmation, but I suspect they will squak at a consumer asking for such an escalation. There are SDK log files that can be pulled and viewed.

What I expect - if there is a bug in the software/firmware it needs to be addressed as not to effect the next pilot that comes across the same scenario. If not a bug, then this points to a hardware failure, which would be warranty. This was a waypoint flight and I feel that the failure is hardware related and not caused by Map Pilot. DJI has incompetent support staff and processes that are unable to troubleshoot (analyze log files correctly) or escalate properly when an issue gets outside of their content or scripting. Someone at DJI knows how to analyze, with a little luck, maybe it will find that someones inbox. It is obvious that whomever looked at the logs had no clue what they were looking at. DJI needs someone to spot check such cases to keep agents in check on their diganoses. This is typical half-baked support that most never question or don't know to question (now I start to rant on support models) and accept incorrect diagnoses from an agent(s) that have no clue. I know there are other avenues yet to take (BBB, Twitter, etc.) but my experience, it can take a couple of months for a qualified individual to review, and in the end, the company never does anything to correct. They will mark this as a "one off" and it's not worth the cost for them to make changes and chalk this up to one unhappy customer vs. 1000 happy customers that fall within current training/process/guidelines. They'll have an acceptable dis-sat metric and this will fall under that metric and no action made from it. Just how the models work. These support folks can make 1 + 1 = 3 on their business reviews when needed so they appear to be doing their job within the defined metrics they are goaled to. One person trying to push is not going to have an effect. When most customers get to this stage I am at, it's a give up, there is no place to go and not worth the fight any longer. The good thing is, Drones are young, there will be a competitor company that will one day start to take DJI market share, then DJI will take a deeper look, lose customers base, etc. - the cycle swings back and forth.

Personally, I think the best option for pilots is getting something like the State Farm Personal Article Policy (PAP). That way you are covered for such scenarios like this for a minmal cost. That's where I am going next step and not worry about shoddy support from a larger company trying to balance a budget. Support is not a money making business so you really can't blame the company. I'm sure they fix most small and easy items they get called about and this fell in to the 4% bottom box metric that will get zero action (I make an assumption their metric is 4%).

BudWalker I was not stating you are associated with any company, you mis-undertstood. I believe you are spot on with your diagnoses and was pointing to that.