Waypoint mission problem

[shb]

Here I try to show what is happening. So pretend that waypoint 1 high, and waypoint 4 is as low as you want to go. You can see with the curve, that it will actually go lower than waypoint 4.

 

[shb]

Would it violate form rules, if I sent a DJ Wes's explanation, in his post number 51, in an email to DJI?

 

[shb]

Here we go again. (Whitesnake). Just got a new email from DJI.

Dear Steven,

Greetings from DJI North America Technical Support.

We just received the response from our relevant team that they attempted four times to get in touch with you via phone call, but failed.

As for now, since the drone experienced a crash accident, it cannot be solved remotely based on our experiences, it is recommended to send your unit in for further checking, to make it possible, kindly please complete the details that were required in the previous email, then we will be able to provide further guidance for you.

As for the waypoint abnormal function, we will make feedback to our engineers to see if a new version will be located in the future to optimize this and bring you a better experience, we care about how you feel and we may feel the same way, but if you can kindly provide the information in the previous email then a reasonable solution or suggestion will be provided, we always believe all the issues can be resolved or we can provide better suggestions for you.

They can't call me because I'm out of the country, and that's not my number, I think it's my wifes. I guess I can give them my Cambodian number. They still think it's a hardware problem, even though DJ Wes has shown otherwise. Hopefully, with the staffs permission, I can send them his explanation.

 

[mobilehomer]

While I hesitate to get involved with such a toxic person, I think I may know what is going on.

The Mavic 3 waypoint implementation uses a Bezier curve to define the waypoint path where the curve passed through the control points. Consider the mission pictured below if it were rotated into a vertical plane. One can see how when traversing from waypoint 3 and through waypoint 4 that the path continues to reach a lower altitude.

Without seeing any details on the OP's mission, this is only a guess. However, depending on the configuration of the OP's mission, this may explain the apparent drop in height below the lower waypoint.

View attachment 164964

That info is in the manual, isn't it? That's why my earlier post was to RTFM!!!! But he is too stubborn to read!!

 

[shb]

That info is in the manual, isn't it? That's why my earlier post was to RTFM!!!! But he is too stubborn to read!!

If you knew it was in the manual, why didn't you point out the specific passage that applies in this case? I assume you're talking about your post #4? I just read the owners manual RTH section, and it says nothing about this problem. Sorry.

 

[mobilehomer]

If you knew it was in the manual, why didn't you point out the specific passage that applies in this case?

You can't read?

 

[shb]

You can't read?

I just read it, and I would say you're the one that can't read, or understand what he's reading!

 

[mobilehomer]

I just read it, and I would say you're the one that can't read, or understand what he's reading!

Try post #23. Apologies will be accepted.

 

[shb]

Try post #23. Apologies will be accepted.

Ok, but that still, does not specifically say, that the waypoint flight will go below the waypoint altitude, does it?

When they said "the aircraft altitude may decrease during the flight" if they had said "the aircraft altitude may decrease, below the waypoint altitude, during the flight" that would be alerting us to the problem. Maybe that's what they meant, but not exactly what they said. Maybe because they're not native English speakers. So if you think that that alerts pilots to the problem of going below the waypoint altitude fine, but I don't think it does.

 

[GadgetGuy]

Would it violate form rules, if I sent a DJ Wes's explanation, in his post number 51, in an email to DJI?

No, but it wouldn't make sense for DJI to also use Bezier curves for changes in elevation where such curves could lead to a lower elevation than programmed, leading to exactly your result: potential obstacle avoidance leading to RTH and cancellation of the mission. DJ Wes is offering a possibility, but a speculative one that would result in changes of a few feet, not 50 feet in elevation. Also, keep in mind that the drone is still alwayprogrammed to fly through each waypoint, even though the path between two points may be a Bezier curve.

Therefore, if your drone is arriving at the lower waypoint 50 feet lower than programmed, this is not a Bezier curve issue!

 

[GadgetGuy]

Here I try to show what is happening. So pretend that waypoint 1 high, and waypoint 4 is as low as you want to go. You can see with the curve, that it will actually go lower than waypoint 4.

View attachment 165077

See my post #70 above.

 

[GadgetGuy]

Ok, but that still, does not specifically say, that the waypoint flight will go below the waypoint altitude, does it?

When they said "the aircraft altitude may decrease during the flight" if they had said "the aircraft altitude may decrease, below the waypoint altitude, during the flight" that would be alerting us to the problem. Maybe that's what they meant, but not exactly what they said. Maybe because they're not native English speakers. So if you think that that alerts pilots to the problem of going below the waypoint altitude fine, but I don't think it does.

Agreed, and "the aircraft altitude may decrease during the flight" might even be Chinglish for "beware of rising ground level below the aircraft" when programming waypoints because they are always relative to your launch point, and not the ground level below the aircraft! "Aircraft altitude" might mean the distance between the aircraft and the ground below it, while still flying level.

 

[shb]

it wouldn't make sense for DJI to also use Bezier curves

Lots of stuff DJI does dosen't make sense! I was thinking the same thing, that even if you're curving the vertical flight, it's not very smart to go below the altitude of the waypoint. In the future, I will put 3 waypoints where the waypoint flight descends significantly.

 

[GadgetGuy]

Lots of stuff DJI does dosen't make sense! I was thinking the same thing, that even if you're curving the vertical flight, it's not very smart to go below the altitude of the waypoint. In the future, I will put 3 waypoints where the waypoint flight descends significantly.

It still is programmed to fly through every waypoint in 3D, regardless of any Bezier curve, so that cannot explain what you have reported, which is that one waypoint is set at 150 feet and the next at 100 feet, and the drone arrives at the second waypoint 50 feet lower than the set 100 feet. That is verbatim what you have reported.

 

[Skyscope]

Lots of stuff DJI does dosen't make sense!

Not for everybody for sure.

I was thinking the same thing, that even if you're curving the vertical flight, it's not very smart to go below the altitude of the waypoint.

Do some homework, that's the way with interpolation of waypoints. There is a choice to make, and DJI decided to plan the mission with a path through the waypoints with Bézier interpolation, with the effects @DJI Wes described in #51.

The alternative would be "The Litchi way" with B-spline interpolation, where the drone never exceeds the imaginary boundary of straight lines between waypoints, but where a curved path may never hits any of them, which means for interpolation, the drone flys curves tighter as planned.
Some would say, that's not very smart because a path shouldn't be flown tighter than projected with waypoints, and I personally agree.

But whatever the alternative, in the end you just have to be smart enough yourself to take the interpolated path into account, instead of blaming others.

 

[GadgetGuy]

Not for everybody for sure.



Do some homework, that's the way with interpolation of waypoints. There is a choice to make, and DJI decided to plan the mission with a path through the waypoints with Bézier interpolation, with the effects @DJI Wes described in #51.

The alternative would be "The Litchi way" with B-spline interpolation, where the drone never exceeds the imaginary boundary of straight lines between waypoints, but where a curved path may never hits any of them, which means for interpolation, the drone flys curves tighter as planned.
Some would say, that's not very smart because a path shouldn't be flown tighter than projected with waypoints, and I personally agree.

But whatever the alternative, in the end you just have to be smart enough yourself to take the interpolated path into account, instead of blaming others.

If you reread Post #51, it only hypothesizes that the Bezier curve also works in 3D, which I find unlikely, but, in any event, that cannot explain what the OP is reporting, which is that the mission is not flying through the waypoints in 3D, but arriving at them 50 feet lower than set. He sets waypoint 1 at 150' and waypoint 2 at 100’, but the drone arrives at waypoint 2 at 50’, 50’ below the set elevation. Assuming that is an accurate statement, Bezier curves between waypoints are not relevant, even if they take place vertically, which I doubt. The mission is still supposed to fly through each waypoint in 3D space, within the accuracy of it's positioning system.

 

[Skyscope]

the OP is reporting, which is that the mission is not flying through the waypoints in 3D, but arriving at them 50 feet lower than set. He sets waypoint 1 at 150' and waypoint 2 at 100’, but the drone arrives at waypoint 2 at 50’, 50’ below the set elevation.

Where did he describe that? Not in #16.

But regardless, I do not speculate on such claims. When it comes to crashes, I only get really involved if the flight log can be viewed at Airdata, for example.
Experience has shown that the problem is usually less with the drone than with the person on the sticks.
And to rule that out, you have to see the original flight log at least, and in this case the Mission Plan too. Otherwise it's all just speculation.

 

[Skyscope]

If you reread Post #51, it only hypothesizes that the Bezier curve also works in 3D, which I find unlikely…

Tested that today, and it does the same in Z-Axis as expected.

Straight path of around 400m lateral distance, 4 Waypoints. The 1st at 50m altitude, 2nd and 3rd at 15m, the 4th at 50m again. The drone went through all waypoints, and down to 11m between waypoint 2 and 3.

 

[GadgetGuy]

Where did he describe that? Not in #16.

But regardless, I do not speculate on such claims. When it comes to crashes, I only get really involved if the flight log can be viewed at Airdata, for example.
Experience has shown that the problem is usually less with the drone than with the person on the sticks.
And to rule that out, you have to see the original flight log at least, and in this case the Mission Plan too. Otherwise it's all just speculation.

Post #5, but to your point, post #16 was rather snarky!
He has not replied to my most recent requests for clarification or verification, so maybe it is all much ado about nothing!

 

[GadgetGuy]

Tested that today, and it does the same in Z-Axis as expected.

Straight path of around 400m lateral distance, 4 Waypoints. The 1st at 50m altitude, 2nd and 3rd at 15m, the 4th at 50m again. The drone went through all waypoints, and down to 11m between waypoint 2 and 3.

Very well. Thanks for converting the speculation into confirmation and disproving my doubts. That could also explain the DJI warning that the aircraft may fly lower than set.

However, the small difference in elevation at the lowest point of 4m in no way can explain the OP's initial claim in post #5. He also claimed that it was overshooting the descent at the waypoint by a full 50’! 150’ to 100’ and arriving at 50’ below the set elevation. We all know that the missions still fly through each waypoint in 3D space, regardless of any rounded Bezier curves while traveling between them in all three axes. He's certainly losing all credibility now, but has prompted an interesting, informative dialogue among us!