Mavic Mini Fly RTH set, triggered by loss of signal, but lands quickly instead

[PeP]

(Help with forensic analysis requested)

The MM was sent on a planned 700 m straight line flight, flat coastal area with scattered apartment blocks 10 – 15 meters height. Wind forecast at 20 meter above ground: 4-5 meter/sec as per UAVForecast (tailwind going out). Flight height above terrain: 30 – 31 meter.

Radio signal was 100% until 490 meters out – suddenly drop to 0%, never to come back. After 30 seconds, RTH auto activated. (The above from the RC flight log file, read via Airdata.com)

Subsequent events: From recorded video onboard: MM rises to RTH altitude: 35 meters and on its return path to RTH Home location. However, after approx. 30 seconds, MM descends and lands, approx. 50 – 100 meters from farthest point of flight (i.e. approx.. 440 – 390 meter from Home point).

Luckily, “Find my Drone” worked nicely and after climbing a locked iron gate, the MM was retrieved from some bushes, unscathed.


Question: Why did the Drone not complete RTH to the starting point ?

I believe I have an obvious answer to a 2.nd question: “What made the radio signal drop from straight 100% to 0% in less than 0.5 seconds, never to be regained while in the air ?”

• Studying video footage (see below), I believe that I lost line of sight, after MM went further away from RC and into the radio shadow of buildings, not so tall, but closer to the RC.

Evidence available:

• RC Flight log file (which was uploaded to Airdata.com )
• Airdata.com HD Flight Player screenprint of last part of RC Flight log
• On-board flight logs. As I understand sar104’s thorough analysis guide, this is not of much help outside dji because of encryption of the MM file (?). Available upon request
• Flight video. This has been truncated to time of landing and downgraded from 1080p/60fps to 720p/30fps and is available here (159MB):
• Failed RTH
• Flight log and video from previous flight, immediately preceding this flight (up and down the beach), and during which there were no signs of malfunction. Available upon request.

Looking forward to your assessments.

 

[Deleted member 94047]

(Help with forensic analysis requested)

The MM was sent on a planned 700 m straight line flight, flat coastal area with scattered apartment blocks 10 – 15 meters height. Wind forecast at 20 meter above ground: 4-5 meter/sec as per UAVForecast (tailwind going out). Flight height above terrain: 30 – 31 meter.

Radio signal was 100% until 490 meters out – suddenly drop to 0%, never to come back. After 30 seconds, RTH auto activated. (The above from the RC flight log file, read via Airdata.com)

Subsequent events: From recorded video onboard: MM rises to RTH altitude: 35 meters and on its return path to RTH Home location. However, after approx. 30 seconds, MM descends and lands, approx. 50 – 100 meters from farthest point of flight (i.e. approx.. 440 – 390 meter from Home point).

Luckily, “Find my Drone” worked nicely and after climbing a locked iron gate, the MM was retrieved from some bushes, unscathed.


Question: Why did the Drone not complete RTH to the starting point ?

I believe I have an obvious answer to a 2.nd question: “What made the radio signal drop from straight 100% to 0% in less than 0.5 seconds, never to be regained while in the air ?”

• Studying video footage (see below), I believe that I lost line of sight, after MM went further away from RC and into the radio shadow of buildings, not so tall, but closer to the RC.

Evidence available:

• RC Flight log file (which was uploaded to Airdata.com )
• Airdata.com HD Flight Player screenprint of last part of RC Flight log
• On-board flight logs. As I understand sar104’s thorough analysis guide, this is not of much help outside dji because of encryption of the MM file (?). Available upon request
• Flight video. This has been truncated to time of landing and downgraded from 1080p/60fps to 720p/30fps and is available here (159MB):
• Failed RTH
• Flight log and video from previous flight, immediately preceding this flight (up and down the beach), and during which there were no signs of malfunction. Available upon request.

Looking forward to your assessments.

From what I can see in the txt file. AC looses signal at 509 seconds. Failsafe action was set to Go Home and the AC switches to GoHome mode as can be seen from the OSD:flycState.



AC re-acquires RC signal at 528s, the flycState continues to be in GoHome mode till the end of the log. However, the failsafe action (MC_PARAM:failsafeAction) changes from GoHome to Hover at 53s.


I can't say why.

 

[PeP]

Thank you, Doppler for your analysis.

I believe I understand why there was loss of signal, as hinted in my problem presentation. Now I will remember, that flying 30 meters above ground and amongst buildings, only 15 meters high, avoids crashes, but not a loss of signal situation, if the operator is at ground level, and the drone flies past and behind building.

After my first post, I read a Fly-away intense discussion, with tempers high, almost to Italian levels, and I realised that my situation could well be explained as follows: Assuming only my downlink transmission were gone, but not my uplink, and waiting for Auto RTH to save the day, when no drone surfaced, I could well have hit the RTH button "to be on the safe side", not realising that I thereby might have cancelled RTH. The video however shows, that the drone has not hovered in place, but has descended directly from RTH altitude to ground landing in less than 30 seconds.

Any thoughts if this sounds probable?

 

[Zelmani]

Not sure if this is of any significance, but why does flight time start with 357 seconds? In the video, two radio towers can be seen on the left and right in front of you. Again not sure, if this might be of any relevance. Could they cause a sudden loss of GPS? What's expected to happen when a GPS loss occurs during RTH an no connection with RC?
At the end, it looks much like one the many uncommanded descents of the Mini, except that this time it was during RTH.

 

[Meta4]

Could they cause a sudden loss of GPS?

No

At the end, it looks much like one the many uncommanded descents of the Mini.

No

 

[Deleted member 94047]

I could well have hit the RTH button "to be on the safe side", not realising that I thereby might have cancelled RTH.

The log shows that the "RC:goHome" parameter is 0 for the entire duration of the log, meaning you did not press the RTH button on the remote. No inputs are registered from the remote after the AC lost connection at 509s which means you did not stop the RTH. But I am at a loss to explain why your AC did not complete the RTH. But I am sure someone else can and will.

PS: About the tempers, you have to take the good with the bad. The flyaway section is one heck of a place to learn; just filter through all the innuendo and the occasional bravado.

 

[Meta4]

AC re-acquires RC signal at 528s, the flycState continues to be in GoHome mode till the end of the log. However, the failsafe action (MC_PARAM:failsafeAction) changes from GoHome to Hover at 53s.

The Failsafe Action shown in the Verbose CSV doesn't change at all.

Luckily, “Find my Drone” worked nicely and after climbing a locked iron gate, the MM was retrieved from some bushes, unscathed.

Where was the drone found?
How far from where signal was lost?

 

[PeP]

@Zelmani: The flight starts at 357 sec.s , because my previous flight stopped at 5min57 sec, And I did not stop propellers, but let it hover. So really a continuation of previous flight, but new log started (Same RTH Home point set). The radio towers (mobile masts) you see are really 300 - 400 meters further away from where I lost contact. I have been flying in this area several days with no observed problems from them.

@Doppler: - The log is from the RC. The AC may have more useful information from after contact was lost, but am I correct in understanding that they cannot be decoded for a Mavic Mini?

 

[PeP]

@Meta4: On the screen print route map, I have marked the landing site with a red star. It was close to point of lost signal, - the drone incl video footage was found unharmed and the footage shows that after hovering at 30 meters for some 15 - 30 sec.s, it rises to 35 meter, starts RTH, but after few seconds initiates landing.

 

[Deleted member 94047]

The Failsafe Action shown in the Verbose CSV doesn't change at all.

CsvView does show it changed though (compare the entries in the "MC_PARAM:failsafeAction" field in the two screen grabs in post #2)

The log is from the RC. The AC may have more useful information from after contact was lost, but am I correct in understanding that they cannot be decoded for a Mavic Mini?

That is also my understanding. If you can find the corresponding DAT file, it may (or may not) shed further light on the incident. See this post for instructions on how to access it. BTW, the txt file does not show the landing. It ends at 536s with the AC at a height of 32.8m, still in a GoHome state (although the failsafe actions changes from RTH to Hover) and apparently decending from the set RTH height of 35m. Did the AC lose connection again? Did you power down the remote? Can you think of anything that might account for the sudden end of the log?



Also, I think it is time to call upon @sar104.

 

[slup]

Yeah ... seeing the failsafe action change to Hover also. This can happen for the Mini if the GPS signal becomes weak according to the manual ... but this doesn't seem to be the case here with 17-18 satellites.



But both from looking at the video clip & reviewing the graphs from the log this COULD again be that "uncommanded descent" we have seen previously ...

At RTH the AC turns, ascending & start to pick up heading speed ... then suddenly it pitches up to +15 degrees (typical for the uncommanding descents) & then the Z downward speed picks up. Just before the AC shows pitch/roll instability... it drifts slightly in the wind direction, which isn't strong, to the right & descending, once below the houses the side movement stops & it only descend.

I think that the DAT log can give us a bit more to look at --> Flight log retrieval and analysis guide (under point 3.)

 

[PeP]

@Meta4: On the screen print route map, I have marked the landing site with a red star. It was close to point of lost signal, - the drone incl video footage was found unharmed and the footage shows that after hovering at 30 meters for some 15 - 30 sec.s, it rises to 35 meter, starts RTH, but after few seconds initiates landing.

 

[PeP]

@slup & @sar104 (via other thread): Thanks for your continuing contributions. At present, I feel that I do not have a satisfactory answer to my question (Post 1), so your and other's comments still very welcome.

I found the relevant .DAT file from my mobile unit, and I enclose a link to it here: 2020-02-05_06-58-04_FLY042.DAT

Two notes: 1) The Dat file contains one previous flight, from T=0 too T=359 sec, and then the fly-away flight from T=359 to landing at approx T=715 sec.
2) I believe that after the 100% loss of signal and Auto RTH, the drone has not communicated with the RC. So the decisive period is probably not in
the linked .DAT file, which is from the RC/mobile, not the encrypted on board the AC.

Again, anybody who has pitched or will pitch in with answers, - taking your time is much appreciated.

 

[slup]

Have seen this behavior before related to "uncommanded descent" events ...

Just before the AC sees the pitch&roll instability the rear motors are more or less maxed out, the pitch angle is around 15 degrees down which looks reasonable. Then suddenly the AC pitch up to 15 degrees goes down again to just a couple degrees down which doesn't give any forward thrust. At the same time with the rear motors maxed the fronts decrease the revs & the descent speed picks up.

A mystery ... anything you can see @sar104 ?

 

[sar104]

Have seen this behavior before related to "uncommanded descent" events ...

Just before the AC sees the pitch&roll instability the rear motors are more or less maxed out, the pitch angle is around 15 degrees down which looks reasonable. Then suddenly the AC pitch up to 15 degrees goes down again to just a couple degrees down which doesn't give any forward thrust. At the same time with the rear motors maxed the fronts decrease the revs & the descent speed picks up.

A mystery ... anything you can see @sar104 ?

View attachment 93427

I'll take a look this evening.

 

[PeP]

After playing around with the RC logs and CSVView, I get the following hypothesis for why the MM aborts RTH and lands where it is instead:

RTH height is set to 35 meters, to which height the MM ascends. The wind velocity in this height is 5.5 to 6 m/sec, head on. The MM on RTH can move at 8 m/sec, so effectively (8 - 5.5 to 6) = 2 - 2.5 m/sec ground speed. The drone is 500 meters away, thus will take 200 to 250 sec.s to get back. Battery level is approx 55%, may not be sufficient to reach Home with acceptable battery minimum reserve (?). So, landing right away.

There is (at least) one flaw in the logic: How can the RTH logic know if MM will continue to battle against this wind, all the way back.

Looking forward to input from experienced log-wrestlers.

 

[Deleted member 94047]

PeP, I can tell you with absolute confidence that your hypothesis doesn't hold water. The battery reserve calculations the FC makes during flight are intended to either warn you to RTH or trigger RTH. Once the AC has entered GoHome mode, the only battery related factor that would trigger a landing is if the battery reaches critical level. Until that happens, the AC will continue flying (or attempting to fly) to the home point; it does not anticipate how long it will take it to go there and just give up and land if it thinks it won't get there. Just wait for sar104 to look at the flight data.

 

[sar104]

Have seen this behavior before related to "uncommanded descent" events ...

Just before the AC sees the pitch&roll instability the rear motors are more or less maxed out, the pitch angle is around 15 degrees down which looks reasonable. Then suddenly the AC pitch up to 15 degrees goes down again to just a couple degrees down which doesn't give any forward thrust. At the same time with the rear motors maxed the fronts decrease the revs & the descent speed picks up.

A mystery ... anything you can see @sar104 ?

There's no definitive evidence, but it looks a bit like another Mini tilt issue. Firstly, the landing wasn't triggered by the smart battery:



This was the second flight on that charge, but it didn't approach either the RTH or the autoland levels.

In terms of the actual behavior, the aircraft entered RTH and went through the align, climb and into the cruise phase. Then it just gave up and landed:



The cruise phase led to unusual tilt excursions as it accelerated, and then it slowed down again and started descending.

 

[slup]

There's no definitive evidence, but it looks a bit like another Mini tilt issue. Firstly, the landing wasn't triggered by the smart battery:

View attachment 93509

This was the second flight on that charge, but it didn't approach either the RTH or the autoland levels.

In terms of the actual behavior, the aircraft entered RTH and went through the align, climb and into the cruise phase. Then it just gave up and landed:

View attachment 93510

The cruise phase led to unusual tilt excursions as it accelerated, and then it slowed down again and started descending.

Yeah, pretty much what I saw also ... the peculiar in this is that the rear motors seems to be pretty much maxed when cruise phase starts with a good -15 degree pitch angle, then the up pitch occurs ... then with still maxed rears the fronts lower the revs but the pitch angle anyway remains pretty horizontal & the drop in height follows.

As said ... a mystery

 

[new2mavic]

I am convinced that there was some problem with the prop / motor of your Mini thorughout the two flights logged in the .DAT file provided in post #13 of this thread. This graph shows the situation at the beginning of the first flight :



At 12 sec , the craft was just hovering at 1.9m. There was some mild wind so the craft needed to tilt slightly ( pitch and roll angle < 4 degree ) to hold it's position. It can be seen that the RBack motor was already running at it's limit ( 15k ~ 16k for Mini ). If the craft is commanded to perform maneuvers requiring more thrust from this motor, it will not be able to fulfil the demand. The RTH action in your second flight is one such maneuver because the craft needed to fly against the wind at the maxiumum ground speed of P mode ( 8 m/s ).

It is possible that there is an untold logic in the firmware of the flight controller that if the craft fails to acheive the maximum allowed tilting angle ( 20 degrees ) or the target ground speed of 8 m/s while maintaining the RTH altitute, the craft will just descend on the spot.

I would suggest you to check the props of the motors at the back. My Mini also has the motors at the back spinning a bit faster when hovering and it was found to be caused by the slightly smaller pitch of the prop attached to those motors. See this thread for further details : Mavic Mini prop has unstable shape ?