Execution Delay
Tried my GPS Test Mission again today.
I kept getting Execution Delay errors.
In this first test, I tried to resume. Dronelink instead began at the Approach point? But again, I got the Execution Delay error.
So in the field, I loaded the mission, cleared the first marker which has all the camera commands, and set only Camera Stop Capture and Camera Mode (Video).
That got me through the original failure, and it began filming.
And the actual footage captured by the Mavic Air before the mission failed.
All this video is completely raw -- sorry if it's boring.
Looking forward to feedback.
Comments
11 comments
The execution delay error means that the device is not able to run the flight controller fast enough for reliable and safe results. Are you using an older device or is the device overheating?
It is an iPhone 6S Plus. I wouldn't call it old, as such.
It doesn't get hot, just warm as usual whenever I use it intensively.
I make a point to kill everything before flying, though I did have the screen and audio being recorded for reporting purposes.
I guess I'll try this once more tomorrow without the screen recording in the background, and see how it goes.
Technically the 6S is a 4 year old device, which from a processor stand-point is ancient. We are doing our best to support lower-end devices (especially because of Android), but in the end there may be trade-offs (like not being able to do screen-recording at the same time).
Yeah, that's fair.
I'll put an iPhoneX on my "needs list". Maybe for Christmas.
Made a run this evening... I can't see the video very well, but it ran the entire plan successfully.
Did you change anything or just get lucky?
I didn't record the phone's screen. That's gotta be a pretty significant processor pig, so as we discussed, I nixed that and tried the plan.
It was dark -- I'd just finished doing my night shot of the lights of the mosque, etc -- so I can't see much in the videos, but the entire course played out.
There was one frightening moment... The drone was flying directly away from me. The Air doesn't have the controller range of the Pro, and this is a noisy environment -- seems you can't swing a cat without hitting a cell tower -- so I typically only get around 400m - 500m range for the controller.
So... I was watching it on the map, and the icon flew past the turn, and kept going, and going.... out over the river, headed towards the President's Palace.
I was on the verge of pressing the RTH, when suddenly the icon jumped back onto the course, apparently having made the turn after all.
-----
I have a point of clarification I'd like to raise. I'd like to understand...
The heavy processor demands + RTH if the controller signal is lost + "built from the ground up" controller......
This equals the entire controlling intelligence stays on my device, right?
I notice that, with Litchi, it uploads the mission to the drone, and if I lose connection with the drone during a mission, it will continue the mission regardless. I've taken that to mean that Litchi loads a series of commands into the drone which the on-board systems process and execute.
Whereas Dronelink is doing that processing entirely on my device, then sending flight and other commands to the drone exactly as how it would happen if I was flying manually. Hence the reason the mission will abort if signal is lost -- the drone doesn't know what to do.
Is that a reasonable conclusion?
Are you saying the drone icon on the map did not accurately reflect the real location of the drone, or that it eventually correctly course after flying off course for a period of time?
Dronelink is unique in that it runs all of the flight control from the device like a true ground station, which is why the drone will RTH on signal loss.
The icon did not reflect the real location. Either my iPhone lagging, or possibly a range issue.
This park I fly a lot in is great in many regards, but is "noisy". I often have video loss at 300-400m, and controller loss about 500-600m, or less if trees, a hill, or structure get in the way. That happens of course.
I shouldn't have mentioned it. This is Beta testing... I should be more informative, less narrative. Sorry.
No apologies needed, keep the feedback coming!
Interesting - just got this error today running another test with my Smart Controller. We may be pushing the SRC's processing capabilities to the limit? One thing I need to mention: I did hover my M2P using GO to set all the camera parameters first, then switched to the dronelink exec without explicitly killing GO. I wonder if that was the real issue.
Please sign in to leave a comment.