Airsense Not Working - DJI Inadequate Response

[ddodell]

One of the reasons I upgraded from my Mavic 2 Pro to the Mavic 3 ( to be honest besides wanting to have the latest version) ... was Airsense. I fly in some areas that have some low flying aircraft, and while I totally understand my responsibilities in avoiding manned flights, Airsense was a tool I wanted to help me meet those responsibilities.

The implementation has never worked well, first with the notifications showing off skewed in the app, to the January firmware which basically killed it. I normally fly in the desert north of Phoenix which also is the beginning of the flight path into a local heavy traffic airport. Using other ADS-B applications, flights are normally 3 to 4 thousand feet above me, but local traffic will drop to 400 to 1200 feet traversing the area, including medical helicopters etc.

Airsense for myself and many others simply is not working in any way. I have not seen a single notification since the January firmware. Today I decided to contact DJI Support ... to be honest, the person didn't know the difference between Airsense and Obstacle Avoidance.

Bottom line, how do we get a $Billion dollar corporation to recognize there is a problem and fix it?

Here is the thread ...

 

[Atkas]

I'll be devil's advocate here. Since the new firmware yesterday which mainly focused on sat acquisition, my Airsense 'seems' to be working correctly. Knowing it was a previous problem, I monitored and compared with my ADS-B receiver (RasPi running 1090) and I was getting correct indications for the 5 - 10 minutes I monitored. Aircraft that weren't transmitting ADS-B obviously didn't/won't show up.
I hope you get yours sorted

 

[Phantomrain.org]

One thing I have noticed, and this goes with many other companies as well is that when there is a New Product Release in Development all there focus goes there first .

Such as the release of the new Enterprise D30 /30T , This seems to be where all there real focus is at and spread the IT team a little thin towards the M3 .

It also explains why the M3 was crippled and limited the way that it is.

Now that the release of the Enterprise is over I think we will see some DJI focus back on the issues but the problem is there are a lot of issues that need help.

I am sure they know of the issues but did not have the man power to address them as they should have been.

Even if I wanted to get the New Enterprise i have little faith that it is Polished

The Mini 3 will be the Next Focus and after that release they might swing back there focus to the M3 . As of now there a hot mess throwing everything out hoping something Sticks because the M3 has fallen some and gotten kind of messy.

Oval Lens - Round lens
Old GPS chip New GPS chip
Quick GSP lock - Long GPS lock
Air Sense Working _ Not working

Phantomrain.org
Gear to fly in the Rain.

 

[ZoneAerial]

One thing I have noticed, and this goes with many other companies as well is that when there is a New Product Release in Development all there focus goes there first .

Such as the release of the new Enterprise D30 /30T , This seems to be where all there real focus is at and spread the IT team a little thin towards the M3 .

It also explains why the M3 was crippled and limited the way that it is.

Now that the release of the Enterprise is over I think we will see some DJI focus back on the issues but the problem is there are a lot of issues that need help.

I am sure they know of the issues but did not have the man power to address them as they should have been.

Even if I wanted to get the New Enterprise i have little faith that it is Polished

The Mini 3 will be the Next Focus and after that release they might swing back there focus to the M3 . As of now there a hot mess throwing everything out hoping something Sticks because the M3 has fallen some and gotten kind of messy.

Oval Lens - Round lens
Old GPS chip New GPS chip
Quick GSP lock - Long GPS lock
Air Sense Working _ Not working

Phantomrain.org
Gear to fly in the Rain.

Did they switch GPS chips in the M3? What do you mean by “old” and “new”?

 

[Phantomrain.org]

Did they switch GPS chips in the M3? What do you mean by “old” and “new”?

This is why I say messy : But here is a thread that was started
Despite this , there was a firmware that did work well with the GPS lock so this may or may not have bearing.

Mavic 3 GNSS chip identified

Popped the cover off the GNSS chip on a Mavic 3 GPS replacement module, and revealed it's a uBlox M8030-KA chip: Previous hardware teardowns exposing chips have shown a slightly different variant of this chip - M8030-KT - in use for at least the Mavic Pro, Mavic 2 Pro, Mavic Mini, Mavic Air...

mavicpilots.com

Phantomrain.org
Gear to fly in the Rain.

 

[ddodell]

I'll be devil's advocate here. Since the new firmware yesterday which mainly focused on sat acquisition, my Airsense 'seems' to be working correctly. Knowing it was a previous problem, I monitored and compared with my ADS-B receiver (RasPi running 1090) and I was getting correct indications for the 5 - 10 minutes I monitored. Aircraft that weren't transmitting ADS-B obviously didn't/won't show up.
I hope you get yours sorted

I haven't flown yet this week ... I hope what you say is true, that the new firmware fixed some of the Airsense issues.

 

[Richard R]

Interesting @ddodell - I see that @Atkas advises that air sense is working. I have RC Pro but have not tested yet with latest firmware. This leads me to believe that the issue may be with the controller?. What controllers are you both using?

 

[ZoneAerial]

This is why I say messy : But here is a thread that was started
Despite this , there was a firmware that did work well with the GPS lock so this may or may not have bearing.

Mavic 3 GNSS chip identified

Popped the cover off the GNSS chip on a Mavic 3 GPS replacement module, and revealed it's a uBlox M8030-KA chip: Previous hardware teardowns exposing chips have shown a slightly different variant of this chip - M8030-KT - in use for at least the Mavic Pro, Mavic 2 Pro, Mavic Mini, Mavic Air...

mavicpilots.com

Phantomrain.org
Gear to fly in the Rain.

Looks like it's getting a bit more clear:

Prompt information - DJI FORUM

,DJI FORUM

forum.dji.com

 

[ddodell]

Interesting @ddodell - I see that @Atkas advises that air sense is working. I have RC Pro but have not tested yet with latest firmware. This leads me to believe that the issue may be with the controller?. What controllers are you both using?

I am using the RC Pro too ... but won't be able to test till tomorrow. Work keeps getting in the way <grin>

 

[MS Coast]

Looks like it's getting a bit more clear:

Prompt information - DJI FORUM

,DJI FORUM

forum.dji.com

The author of that DJI Forum post blames the problem on two things: 1) an "automotive" grade GPS chip that has been used in some Mavic 3s and has less memory and 2) lack of memory to store the ephemeris data from GPS satellites (position and other data that facilitates use of the satellite to calculate a position) because of the addition of FocusTrack, Hyperlapse, and MasterShots, which use the GPS position.

There are some plausible aspects of the post, but it's not clear why the memory requirements of the intelligent flight modes would increase the time to download the ephemeris data. The reports here don't support the idea that using intelligent flight modes caused subsequent poor GPS lock times. Reports talk about consistently poor lock times, not good one day and bad another.

 

[ZoneAerial]

The author of that DJI Forum post blames the problem on two things: 1) an "automotive" grade GPS chip that has been used in some Mavic 3s and has less memory and 2) lack of memory to store the ephemeris data from GPS satellites (position and other data that facilitates use of the satellite to calculate a position) because of the addition of FocusTrack, Hyperlapse, and MasterShots, which use the GPS position.

There are some plausible aspects of the post, but it's not clear why the memory requirements of the intelligent flight modes would increase the time to download the ephemeris data. The reports here don't support the idea that using intelligent flight modes caused subsequent poor GPS lock times. Reports talk about consistently poor lock times, not good one day and bad another.

The reports do support that. Because GPS lock times were just fine before they implemented the firmware which included all the intelligent flight modes.

 

[GadgetGuy]

The reports do support that. Because GPS lock times were just fine before they implemented the firmware which included all the intelligent flight modes.

True. Support being the operative word. It's a possibility based upon a correlation. Clearly, something in the December update caused the problem, as no one had this issue previously, and a current rollback to the prior firmware before December on all hardware restores the quick acquisition times. This is as good an explanation as any! Absent any other, I'm going with it for now.

 

[ZoneAerial]

True. Support being the operative word. It's a possibility based upon a correlation. Clearly, something in the December update caused the problem, as no one had this issue previously, and a current rollback to the prior firmware before December on all hardware restores the quick acquisition times. This is as good an explanation as any! Absent any other, I'm going with it for now.

That DJI hasn't been able to do anything in three months to fix the problems caused by the first firmware update should tell you that they have a real problem that's not easily solved.
It all fits perfectly.
Why do you think that a company as big as DJI hasn't been able to solve the GPS issue in 3 months?
Do you have a better explanation?

 

[Suren]

True. Support being the operative word. It's a possibility based upon a correlation. Clearly, something in the December update caused the problem, as no one had this issue previously, and a current rollback to the prior firmware before December on all hardware restores the quick acquisition times. This is as good an explanation as any! Absent any other, I'm going with it for now.

I had did a rollback via drone hacks app to FW 0300 which is prior to the December update and my GPS lock was like 30 seconds. It was super fast.

 

[GadgetGuy]

That DJI hasn't been able to do anything in three months to fix the problems caused by the first firmware update should tell you that they have a real problem that's not easily solved.
It all fits perfectly.
Why do you think that a company as big as DJI hasn't been able to solve the GPS issue in 3 months?
Do you have a better explanation?

I don't, so this one seems as plausible as any, and it comes with an air of authenticity. I'm happy my own workarounds of a power cycle at 8 satellites, or immediately ascending to 96 feet, are both working to keep my HP acquisition to under 2 minutes on a cold start.

Worst case scenario, have a spare battery solely for HP acquisition, and swap batteries for a launch on a 100% battery on a warm boot immediately after.

 

[Suren]

Airsense works perfectly so far on AC FW version 0300 and app 1.5.8 - I just did a test and it was perfect. I am waiting for another plane to come pass so I can test the 1.5.8 app with the latest AC FW

 

[ianwood]

This is what I wrote on the DJI forum:

ianwood Posted at 3-25 22:30

Can you provide a link to the other forum where it was found the memory of the KA version to be different? How about a reference to the spec sheet where this difference is identified? If it impacted performance as much as we have seen, surely this difference is material enough to be documented. Also, there was always a noticeable difference between cold and warm start on all versions of the FW. So not sure how that is consistent with this explanation. I don't see how the features that came in the December update would use additional memory in the GPS chip. They might use more memory in the flight controller where navigation routines are carried out for those features but the GPS chip should simply be sending NMEA sentences or similar. I can't see any need to store additional coordinates or other data on the GPS chip itself. Lastly, A-GPS does not seem to be implemented or working in any noticeable capacity on any fimrware version. At least not from the RC Pro (even though RC Pro itself uses A-GPS for its own GPS functions).

We do know for fact the Mavic 3 does use the KA (automotive) version of the uBlox 8030. This much is true. The rest I am struggling to make full sense of it as it doesn't seem to add up on the surface of it. The detail will explain it one way or the other.

 

[ddodell]

In my original post, I asked DJI Support if they could define the specs on the new Airsense. They were happy to share the specs from the manual on obstacle avoidance. My only objective was to have a better idea on when it really should be working. Just got this answer from them ... all they keep doing is promising my safety, but no real definition under what circumstances Airsense should work.

Hannah （DJI Support）

2022年3月26日 上午9:27 [GMT+8]

Dear Customer,

A pleasant to you!

First of all, we would really like to thank you for your patronage in our products.

We do apologize for the inconvenience for the long wait before we could be able to get in touch with you. Please allow me to inform you with regret that we are unable to provide you with the relevant data of Air Sense. But rest assured that we just adjusted the Air Sense function in V01.00.05.00. and we will ensure the safety of the flight.

We are committed to providing you with a great flying experience and thank you for your attention to the performance of DJI products.

Your understanding and patience will be greatly appreciated. Have a great day!

Should you have any other concerns, please don't hesitate to contact us back. We are happy to help you.
Thank you for choosing DJI.
Best Regards,
Hannah
DJI Technical Support

 

[MS Coast]

The reports do support that. Because GPS lock times were just fine before they implemented the firmware which included all the intelligent flight modes.

Coincidence and cause/effect can both provide an explanation.

What's unsupported is the idea that the addition of intelligent flight features causes the memory problem. Do the intelligent flight features use memory in the GPS module immediately on startup? And why would that increase satellite acquisition times?

It's definitely possible that different GPS hardware might be the cause of the problem. But if that's the case, why do some users, presumably with the bad hardware, report that the new firmware release resolved the problem?

 

[EpicFlight]

Folks, you're way too suspicious. I've written code for things like this. It's very technically complex. Having that experience, I'm not suspicious of this explanation at all, and it rings very credible in the details.