I noticed that the Air 3 flies a vertical bezier curve between waypoints with different altitudes, this can lead to unwanted situations (heights).
Example:
WP1 to WP17 H=40m, WP18 H=55m (last waypoint).
Between WP16 & WP17 the drone slowly descents from 40m to 27m then slowly ascents to 40m again at WP17 to get to a height of 55m at WP18.

Hi, fans302fcd3a. Thank you for reaching out. Let me forwrad this to our relevant team for further checking. We'll let you know once we receive an update.

Hi, fans302fcd3a. Thank you for patiently waiting, we now received feedback from our relevant team. As per them, based on the video that you have provided, it is the normal logic. The trajectory height change between two waypoints will be affected by the height of the waypoints before and after the two waypoints. The flight route will curve between waypoints, so the aircraft altitude between waypoints may become lower than the altitudes of the waypoints during the flight. Hope this helps. Thank you for your continued support.

DJI Gamora Posted at 9-9 17:57
Hi, fans302fcd3a. Thank you for patiently waiting, we now received feedback from our relevant team. As per them, based on the video that you have provided, it is the normal logic. The trajectory height change between two waypoints will be affected by the height of the waypoints before and after the two waypoints. The flight route will curve between waypoints, so the aircraft altitude between waypoints may become lower than the altitudes of the waypoints during the flight. Hope this helps. Thank you for your continued support.

Hi DJI Gamora,

Your explanation is what I noticed, so nothing new.
It is indeed 'the normal logic' when you use bezier curvers in the vertical plane.
My point/issue is that you should use another method in the vertical plane, like you do in your SDK's which is used by 3rd party developers like Litchi.

The flown flight plan is a copy/conversion of a Litchi waypoint mission, and when flown with my Phantom 4 does NOT show this unwanted descent between WP16 & WP17.

DJI Gamora Posted at 9-9 17:57
Hi, fans302fcd3a. Thank you for patiently waiting, we now received feedback from our relevant team. As per them, based on the video that you have provided, it is the normal logic. The trajectory height change between two waypoints will be affected by the height of the waypoints before and after the two waypoints. The flight route will curve between waypoints, so the aircraft altitude between waypoints may become lower than the altitudes of the waypoints during the flight. Hope this helps. Thank you for your continued support.

I have experienced the exact same issue - and it IS AN ISSUE, despite the auto-reply from the "relevant teams"!
If you program (as I did, which led me to this forum post) a flight where the 3 first waypoints are set to an altitude of 3 meters, and then sets the altitude for waypoint 4 to 10 meters, the mission fails because the drone tries to bury itself in the ground between waypoint 2 and 3. I can already guess the moderators reply, but do you really not see an error in this "normal logic"?

Clouseau Posted at 11-12 13:14
I have experienced the exact same issue - and it IS AN ISSUE, despite the auto-reply from the "relevant teams"!
If you program (as I did, which led me to this forum post) a flight where the 3 first waypoints are set to an altitude of 3 meters, and then sets the altitude for waypoint 4 to 10 meters, the mission fails because the drone tries to bury itself in the ground between waypoint 2 and 3. I can already guess the moderators reply, but do you really not see an error in this "normal logic"?

Hello, there. Sorry for the late notice and thank you for your patience. While I completely understand that you're experiencing a situation relative to this post, I'd like to ask if you could send us a video of your flight experience. You may upload it through Google Drive or Dropbox and provide the link here. I hope to hear from you soon.

I've now experienced this issue in reverse. I had a mission fail in a 60m altitude zone and the curvature through a waypoint placed at 60m caused the mission to fail. Not a tragedy, but a demonstration that, as pilots, we need to comprehend the 3-dimensional nature of our flight path. Had I left a "buffer" and not placed that waypoint right on the limit I'd have been OK. The nature of Bezier curves is to allow the desired "smooth transition" through direction changes which includes the vertical movement. The flight path that I created tried to fly the drone vertically right through that 60m altitude limit!
When you wrap your mind around the fact that we fly in 3 dimensions, you'll understand that a 2D "birds-eye" representation of your path is inadequate. Waypoints are simply Bezier control points, which require some understanding.

My solution is to place additional waypoints to ensure safe *vertical* transitions. One WP with elevation change and no heading change; the next WP changes the heading, no altitude change. This issue is actually easy to replicate and solve once you know that you're drawing curves in 3D space. IMHO, not a DJI issue but a pilot education issue. However, the waypoint process should be better documented going forward. We're all very glad to have the new waypoints feature.

Same here @djigamora... I have a mission that keeps failing because I hit max altitude (waypoint 1 at 50m altitude, waypoint 2 at 100, waypoint 3 at 100, but the route takes the drone over 120m in between, and waypoint mission fails).

Would it be difficult to make the waypoint mission respect the max height setting, as long as it is below the height actually set on a waypoint? I would understand if the waypoint altitude would be set above the max height, then there should be an error, but not if it is anyway just an internal calculation.

Curved flight paths between points in the horizontal plane make tons of sense and are awesome. Not so for the vertical plane.