If this helps any newbies like myself.......

This has probably come up many times, but I'm new to the forum. I got a little over confident the other day and wanted to get a good river shot. I had good satellite numbers and no hovering issues before or after, so I'm guessing this was due to the bottom sensors winning the tug of war with GPS. I carefully avoided all the branches, but at the 22 sec mark it just started to drift with the current. There were quite a few branches directly downriver, so I'm lucky I had my eyes on the Mini and I was directly behind it. I didn't move the stick in the wrong direction.

Just a little reminder to be careful out there. I almost made an expensive mistake going for the cool shot. Hopefully you won't.

Yes, I've had the same thing happen. Must be the sensors trying to distinguish the water from land or something.

Agree, the current is causing the MM to move with it as the VPS sensors are seeing the current as the land beneath it moving so it moves with it. Good video, maybe just a tad more altitude so as not to confuse VPS might do the trick.

Hi. Thank you for sharing your mission and your insights with regards to this experience. I am happy to know that your DJI Mavic Mini is safe. Flying above water or any reflective surface may cause the vision system to not work properly, please fly with caution when flying above water. For additional information, I will be posting an image for the downward vision sensor. Thank you and safe flying.

Does the VPS have a range setting? Maybe it would be nice to be able to set a distance.

Thanks for sharing, definitely good reminder for new pilots.

Hiya,

Very interesting this!  AFAIK when the GPS is available the craft will use GPS data to stabilize and not the vision sensors, guess i am wrong.

Do you mind to upload this flightlog, love to see if the vison was indeed active and used or not.

link > https://www.phantomhelp.com/LogViewer/Upload/

cheers
JJB

I'm pretty sure, that BOTH are enabled, GPS and VPS. VPS is never deactiveted, unless you cover the sensors. Also I think if the speed over water is high enough, that there would be no problem as CarRamRod has reported. This would a good option in the flyapp ... the option to disable the sensors. I would like to fly into a tight tube, but with VPS no chance
@Wiz33 : There is no range to set, VPS is effective from nearly ground up to 10 meters

JJB* has a point. Seeing the log for that flight would shed a little more light on it. If it was in GPS mode, then the VPS system shouldn't have affected it.

JodyB Posted at 12-18 09:11
JJB* has a point. Seeing the log for that flight would shed a little more light on it. If it was in GPS mode, then the VPS system shouldn't have affected it.

I agree, would be great to see what happened. My guess is that being close to moving water, the DVS thinks foreground is moving.

InspektorGadjet Posted at 12-18 09:18
I agree, would be great to see what happened. My guess is that being close to moving water, the DVS thinks foreground is moving.

That's my thinking too. But Like JJB* pointed out, if the MM was in fact in GPS mode instead of VPS, it should have opted to use the GPS data to hold its position.

It will be interesting to see the logs from this one me thinks!!

I'll try and get the logs this evening. At work right now.

jonny007 Posted at 12-18 09:03
I'm pretty sure, that BOTH are enabled, GPS and VPS. VPS is never deactiveted, unless you cover the sensors. Also I think if the speed over water is high enough, that there would be no problem as CarRamRod has reported. This would a good option in the flyapp ... the option to disable the sensors. I would like to fly into a tight tube, but with VPS no chance
@Wiz33 : There is no range to set, VPS is effective from nearly ground up to 10 meters

true, when not diabled vps is active. But there is a difference being active and if this data is used.
So i am curious to see the log.

Dano99 Posted at 12-18 07:44
Yes, I've had the same thing happen. Must be the sensors trying to distinguish the water from land or something.

what sensors? it's mini.

izotop Posted at 12-18 16:38
what sensors? it's mini.

The mavic mini has two sensors on the bottom back side of It. They serve the mini with a VPS system. They sense it’s position over ground for opti mode.

JJB* Posted at 12-18 08:35
Hiya,

Very interesting this!  AFAIK when the GPS is available the craft will use GPS data to stabilize and not the vision sensors, guess i am wrong.

Interesting looking at the logs. I did have GPS errors right before going out over the river. I'm not 100% sure what I'm looking at though. Here is the link:

https://www.phantomhelp.com/LogViewer/IFVBF84WDRP1DYONZNCF

Turning  off bottom sensors may or may not eliminate danger of going submarining.  Public GPS positioning is not down to centimeters.  Which could lead to drone deciding to relocate itself lower.  Relying on you to quickly command higher.

CarRamRod Posted at 12-18 18:30
Interesting looking at the logs. I did have GPS errors right before going out over the river. I'm not 100% sure what I'm looking at though. Here is the link:

https://www.phantomhelp.com/LogViewer/IFVBF84WDRP1DYONZNCF

The "GPS Position NoMatch" (or "GPS Position No Match") have shown up in other people's flight logs.  Usually without incident.  On some occasions prior to a crash.  DJI hasn't said what message means.

One pilot theorized it was result of GPS satellite signal interference, possibly trees overhead.

Theorizing a couple satellites were lost and different satellites picked up, for brief moment system gets difference in drone's GPS Position until drone's GPS system catches up, making corrections to get GPS lock.

CarRamRod Posted at 12-18 18:30
Interesting looking at the logs. I did have GPS errors right before going out over the river. I'm not 100% sure what I'm looking at though. Here is the link:

https://www.phantomhelp.com/LogViewer/IFVBF84WDRP1DYONZNCF

Thanks for your log.

Interesting log, i will try to explain.
Number of satellites during flight were fluctuating, see chart.
Number of receiving sats and quality of reception is important for hover (no stick input) accurancy.

First part of your flight distance to HomePoint < 5 meter, see chart.
Coloured moving line (no RC input for Pitch and Roll) is from the actual GPS readings, so or craft was stable and gps data is fluctuating or gps data is correct and craft moved along this line. This polar chart is from start to about 11minutes.

I saw your video, did find the above water flight part in the log.
As you can see the VPS height (green bold line) stops measuring/indicating.
So the reference to hold controlled position is lost, thus craft cannot hold position.
I think that because of the less accurancy of GPS signals craft starts to 'move' on his own.

Last chart is the moving line flying over water. I have no clue why the line is 'broken' in the chart, must be a bug in my program. (keeps me busy now...)

So my conclusion is that MM uses its Bottom sensor (when measuring) for controlling the position, if out of range or not measuring GPS takes over.
But need more logs to check if this is really the case.

EDIT  i did check my brand new mini in my house, and in my house the VPS height does indicate active or not, flying from inside my house to outside and GPS became available it changed into GPS hold position, thus bottom sensor measuring but not used for positioning.Thus, i think, that in the above flight position was controlled by GPS.
See chart#6 for my flight indoor > outdoor


cheers
JJB

Good video

JJB* Posted at 12-19 01:36
Thanks for your log.

Interesting log, i will try to explain.

Wow! Thanks for all this info. I appreciate the time you put into this!

Ok, after analysis what would the recommendation be if I fly over water and don't want any irritation from the sensors ? Could you say no deeper than 1 meter, rather fast than slow and beware of hovering close to the surface ?