Autopilot - tips?

[Tenly]

Please sign up for the beta here.

If you've already subscribed to the mailing lists, that form throws an error (still).

Same problem I had a month ago when I tried to join the beta - so I followed that up with an email to support - but still didn't hear anything...

Tell the truth.... Is there really a beta program or not?

 

[Hangar]

Tell the truth.... Is there really a beta program or not?

Apologies if it slipped through the cracks. I promise if you send another email to support it will be sorted out.

 

[pilotpeter]

What data does it load to healthy drones? I just completed a waypoint missing with lcmc. After the mission I noticed not much of the data was uploaded to healthy drones, it appeared it stored only data from when there was established RC-quad communication. Is this a limitation?

 

[pilotpeter]

I noticed the video stopped recording as well after lost comm. is that correct or one of my incorrect settings possibly?

 

[Hangar]

What data does it load to healthy drones? I just completed a waypoint missing with lcmc. After the mission I noticed not much of the data was uploaded to healthy drones, it appeared it stored only data from when there was established RC-quad communication. Is this a limitation?

The flight recorder just records the telemetry that is received from the aircraft while it is connected. To the extent DJI does their own logging on the aircraft itself, you may have to use their files in this case.

 

[Jim North]

I recently upgraded to Autopilot 3.4 along with the following:
DJI Go app 2.8
P3A 1.08
RC 1.6.0

Everything works except the video feed which is now broken up. Meanwhile the video feed in DJI Go works fine.
Any suggestions? I love my Autopilot and use it all the time.

 

[Hangar]

I recently upgraded to Autopilot 3.4 along with the following:
DJI Go app 2.8
P3A 1.08
RC 1.6.0

Everything works except the video feed which is now broken up. Meanwhile the video feed in DJI Go works fine.
Any suggestions? I love my Autopilot and use it all the time.

Did you upgrade everything all at once, or piecemeal? If piecemeal, do you know which update started the video feed issue?

 

[Jim North]

Did you upgrade everything all at once, or piecemeal? If piecemeal, do you know which update started the video feed issue?

Autopilot downloaded automatically first on my iPad Mini. Then I did the aircraft followed by DJI Go and last the RC.

No I don't know when the video feed broke up - I didn't fly until all updates were complete. All updates were within 24 hours.

 

[Hangar]

Can you do a screen recording and send a link to [email protected]

 

[Jim North]

Can you do a screen recording and send a link to [email protected]

OK I'll try that - thanks for your help

 

[Jim North]

Can you do a screen recording and send a link to [email protected]

Hey my Autopilot is back to normal and I made no changes whatsoever.
I made a flight to get a screen shot to show support the video feed breaking up and now the video is back as good as always.
If anyone else has the video feed go strange while in Autopilot, please post it here so we can figure out what is causing it.
It was odd because the video feed to DJI Go at the same time was good.

Sometimes we are 4 hours back on a wilderness trail and I need to know that Autopilot is going to be 100% reliable.

 

[2nd2non]

It was odd because the video feed to DJI Go at the same time was....

If I follow you correctly, seems you have DJI GO and Autopilot open at the same time. This has proven to unstable at times and use to vary from firmware to firmware depending on whatever DJI was doing. As a rule, I stopped using two apps at same time to avoid misc issues.

 

[Jim North]

If I follow you correctly, seems you have DJI GO and Autopilot open at the same time. This has proven to unstable at times and use to vary from firmware to firmware depending on whatever DJI was doing. As a rule, I stopped using two apps at same time to avoid misc issues.

That's a good point but I always close DJI Go, switch from P to F and open Autopilot.

 

[mortenp]

Hi,

I have been trying out autopilot, specifically the "Cruise" mode. I want to control some aspects of the flight, while autopilot keeps the yaw focused on the r/c location. So I set vertical and pitch/roll to joystick, and yaw strategy to focus, focus strategy to subject, and subject to operator.

I calibrate compass, ensure sufficient GPS on both phantom 4 and iPhone, lift off using remote, and engage. One engaged, the phantom just takes off in one direction, without any input from me, and continues flying away from me until I get worried and stop it, by switching to sport mode and bringing it back. Usually I have stopped it after it has flown about 60-70 m away. Direction has been either to the east or west, and it is not focusing on me at all.

I'm in open areas far from any buildings or antennas, and autopilot shows my location correctly on the map.

Obviously I'm expecting or doing something wrong, I was expecting to be in control of at least some aspects of the flight.

Has anybody used cruise mode with success, and can explain to me which settings work? I'm afraid of using the app now, since I feel as if I have no control at all over it.

Thanks in advance,
morten

edit: It turns out it was probably light winds pushing my phantom around.

 

[Hangar]

I calibrate compass, ensure sufficient GPS on both phantom 4 and iPhone, lift off using remote, and engage. One engaged, the phantom just takes off in one direction, without any input from me, and continues flying away from me until I get worried and stop it, by switching to sport mode and bringing it back. Usually I have stopped it after it has flown about 60-70 m away. Direction has been either to the east or west, and it is not focusing on me at all.

Please email the flight logs, including the ones in the engagements folder to [email protected].

 

[mortenp]

Please email the flight logs, including the ones in the engagements folder to [email protected].

Thank you, I have mailed to you.

morten

 

[Tenly]

I've submitted another feature request via the support link on the autoflightlogic website - or more accurately an update to a feature I requested many weeks ago.

In the hopes of generating at least a couple "me too's" for this feature request, I paste my email below...

<begin paste>
Several weeks ago I posted a suggestion that you should allow users to capture their own image data for use as a backdrop when creating waypoint missions.

Since then, it has come to my attention that such a service already exists in a product/service called mapbox.com and there is at least one other drone automation app (Map Pilot) that allows users to select the source of the background imagery from options like Apple Maps, Apple Street Maps, Google Maps, Open Street View Maps and MapBox maps.

I would request that you drop everything and immediately redirect the focus of every single employee you have into adding at least the MapBox option as an option in Autopilot. I have some pending work to do that requires waypoint flying in an area with very poor Google imagery and require an alternate imagery source.

With the entire company focusing on this, I'm sure you could have these features added to the beta in about a week. Please let me know if your estimates are different from mine....

Okay....so I hope that gave you a chuckle...but in all seriousness, I think this might be a very simple change to make in your software. I would be extremely surprised if MapBox did not have an easy way to move existing products from Google maps to MapBox maps. They probably use identical API's to make it easy for software developers to switch. The only real effort required from your developers will be updating the GUI to allow users to select MapBox as their imagery source.

There may be a cost associated with providing MapBox as an option. I don't know how much or if you would be able to absorb that cost with your current pricing structure but I for one would be willing to pay a few dollars more via an in-app purchase for this capability and I suspect that any other users that need to fly in areas which Google has not updated recently might also be willing to pay extra.

So I was kidding about redirecting the focus of all employees - but in reality, I believe this will be an easy capability to introduce on the engineering side of things and I would ask that you have one of your senior engineers explore this possibility immediately rather than add it to a backlog of feature requests. I don't believe that Litchi currently has this capability so this would be a cool differentiator for your software to have.

As I suspected, they have a guide on their website for migrating your app from Google Maps to MapBox maps using either objective-C or Swift:
Switch from Google Maps to the Mapbox iOS SDK | Mapbox

<end paste>

 

[mortenp]

I've submitted another feature request via the support link on the autoflightlogic website - or more accurately an update to a feature I requested many weeks ago.

In the hopes of generating at least a couple "me too's" for this feature request, I paste my email below...

<begin paste>
Several weeks ago I posted a suggestion that you should allow users to capture their own image data for use as a backdrop when creating waypoint missions.

Since then, it has come to my attention that such a service already exists in a product/service called mapbox.com and there is at least one other drone automation app (Map Pilot) that allows users to select the source of the background imagery from options like Apple Maps, Apple Street Maps, Google Maps, Open Street View Maps and MapBox maps.

I would request that you drop everything and immediately redirect the focus of every single employee you have into adding at least the MapBox option as an option in Autopilot. I have some pending work to do that requires waypoint flying in an area with very poor Google imagery and require an alternate imagery source.

With the entire company focusing on this, I'm sure you could have these features added to the beta in about a week. Please let me know if your estimates are different from mine....

Okay....so I hope that gave you a chuckle...but in all seriousness, I think this might be a very simple change to make in your software. I would be extremely surprised if MapBox did not have an easy way to move existing products from Google maps to MapBox maps. They probably use identical API's to make it easy for software developers to switch. The only real effort required from your developers will be updating the GUI to allow users to select MapBox as their imagery source.

There may be a cost associated with providing MapBox as an option. I don't know how much or if you would be able to absorb that cost with your current pricing structure but I for one would be willing to pay a few dollars more via an in-app purchase for this capability and I suspect that any other users that need to fly in areas which Google has not updated recently might also be willing to pay extra.

So I was kidding about redirecting the focus of all employees - but in reality, I believe this will be an easy capability to introduce on the engineering side of things and I would ask that you have one of your senior engineers explore this possibility immediately rather than add it to a backlog of feature requests. I don't believe that Litchi currently has this capability so this would be a cool differentiator for your software to have.

As I suspected, they have a guide on their website for migrating your app from Google Maps to MapBox maps using either objective-C or Swift:
Switch from Google Maps to the Mapbox iOS SDK | Mapbox

<end paste>

The idea of being able to snap a photo from above your subject, and using that as a mission planning map, is really great. You would be able to put waypoints between trees and other things, even avoiding temporary obstacles.

Good idea!

 

[mortenp]

Please email the flight logs, including the ones in the engagements folder to [email protected].

Thanks, it turns out it was probably light winds pushing my phantom around.

 

[Tenly]

The idea of being able to snap a photo from above your subject, and using that as a mission planning map, is really great. You would be able to put waypoints between trees and other things, even avoiding temporary obstacles.

Good idea!

Thanks! Hopefully they'll see the value in it and get it added as a capability sometime soon. Using MapBox is going to have a cost associated with it - hopefully AutoFlightLogic will not withhold an important feature from us because it cuts into their profits too much. Your point regarding temporary obstacles is a good one and it's one I hadn't thought of. That makes this a good feature to have even in areas where Google provides clear and recent imagery!

As for flying in between trees - even if you have super accurate imagery - your GPS location will vary based on how many satellites you're currently connected to and other factors - so unless it's a pretty big passage between them - I'd still be cautious about trying to fly in between them!