Low pressure systems affecting waypoint missions

[shb]

Today it is cloudy and cool, a low pressure system has moved in. Will my waypoint missions fly lower or higher than normal. Thanks.

 

[GadgetGuy]

The launch point represents the zero elevation of the Waypoint mission. So, unless the barometric pressure at launch changes significantly during the mission, they will fly as you intended.

 

[shb]

The launch point represents the zero elevation of the Waypoint mission. So, unless the barometric pressure at launch changes significantly during the mission, they will fly as you intended.

Thanks for responding. My mission seemed to fly a little low. Doubt the weather would change much in one mile.

 

[slup]

The mission height is dependent & in relation to the launch point... so if you design a mission where the drone flies 10m above a special intended launch point... & then launch the mission from another launch point located 10m below the original, the mission height will be lowered by those 10m all over & the drone will hit ground.

So if your mission seemed to fly a little low... did you launch in another place located a little lower?

*EDIT*
Ah... saw now it was the native M3 waypoint navigation & not Litchi. But it maybe works in a similar way?

 

[shb]

The mission height is dependent & in relation to the launch point... so if you design a mission where the drone flies 10m above a special intended launch point... & then launch the mission from another launch point located 10m below the original, the mission height will be lowered by those 10m all over & the drone will hit ground.

So if your mission seemed to fly a little low... did you launch in another place located a little lower?

*EDIT*
Ah... saw now it was the native M3 waypoint navigation & not Litchi. But it maybe works in a similar way?

No I always launch at the same place. My driveway.

 

[slup]

No I always launch at the same place. My driveway.

How much is "a little low"?

 

[shb]

How much is "a little low"?

I would say 5 or 10 feet but not sure. I'm trying to go as I can without hitting any obstructions. Following the path of a river that is mainly vegetation now.

 

[slup]

I would say 5 or 10 feet but not sure. I'm trying to go as I can without hitting any obstructions. Following the path of a river that is mainly vegetation now.

That's within the margin of the barometric sensor during a flight... wouldn't recommend designing a mission so close to obstacles that 5-10ft matters.

 

[AnzacJack]

if the drone acted the way you are suggesting, the drone could be out by 100s of feet.
The hPa unit that we use down here equates to 30ft every 1 hPa and this could range from 1008 to 1018 on a day to day basis here in Australia (can certainly go higher or lower, but just ball parking the figures)

 

[shb]

That's within the margin of the barometric sensor during a flight... wouldn't recommend designing a mission so close to obstacles that 5-10ft matters.

Yeah, I know I'm pushing it, but with the mavic 3, worst case it will do a RTH. Hopefully! But the lower you go, the better the video looks, within reason.

 

[MS Coast]

Yeah, I know I'm pushing it, but with the mavic 3, worst case it will do a RTH. Hopefully! But the lower you go, the better the video looks, within reason.

Actually, if the drone encounters a thin branch or wire in its path , the best thing that can happed is an RTH (or stop, according to your settings). The more likely thing is that it will fly into the obstacle.

Thin branches without leaves, vines, and wires often aren't detected by the obstacle avoidance sensors. A 5 to 10 foot clearance margin is pretty slim, especially if the trees are growing and the waypoints were set some time ago.

 

[shb]

Actually, if the drone encounters a thin branch or wire in its path , the best thing that can happed is an RTH (or stop, according to your settings). The more likely thing is that it will fly into the obstacle.

Thin branches without leaves, vines, and wires often aren't detected by the obstacle avoidance sensors. A 5 to 10 foot clearance margin is pretty slim, especially if the trees are growing and the waypoints were set some time ago.

Yes, wires are the worst. I did hit a tree with no leaves, but didn't crash. Here, they have these kids flying homemade kites. The strings would definitely not be detected, but they don't fly too high. Two days ago, my mavic 3 detected a lightning rod, maybe 3 inches in diameter. It also made a RTH, when a bird was flying toward it.

 

[shb]

Yesterday I flew my mission, and it successfully completed. So I made some lateral adjustments because it was slightly off to the left in one place. Today, when I went to test it, it came back in 6 minutes. It's a 15 minute mission. So I found out were it ran into a banana tree, and increased the height there 6 ft. Ran it again, and it came back in 8 minutes. It hit a clump of tall foliage. I found out the srt file gives relative and absolute altitude. When I checked to mission, the absolute altitude was 25 feet lower today. I was shocked, and there's no way to automatically raise every waypoints elevator by 25 feet. So this is a big setback. Here's a screen shot of the exact same place on different days.

 

[GadgetGuy]

Yesterday I flew my mission, and it successfully completed. So I made some lateral adjustments because it was slightly off to the left in one place. Today, when I went to test it, it came back in 6 minutes. It's a 15 minute mission. So I found out were it ran into a banana tree, and increased the height there 6 ft. Ran it again, and it came back in 8 minutes. It hit a clump of tall foliage. I found out the srt file gives relative and absolute altitude. When I checked to mission, the absolute altitude was 25 feet lower today. I was shocked, and there's no way to automatically raise every waypoints elevator by 25 feet. So this is a big setback. Here's a screen shot of the exact same place on different days.
View attachment 162140

Not exactly the same place. Longitude in one is 120.58729 and 120.587298 in the other.
The abs_alt is actually the rel_alt because the only elevation DJI can measure is relative to the launch point.
For greater accuracy of the relative elevation changes measured by the barometer, let it warm up to better calibrate itself, by letting it sit for a minute or two, before arming the motors, after turning the battery on, even if the Homepoint has already been acquired. As noted by others, the barometric accuracy can easily be off by 10-25 feet, so add an additional 25 feet to all your waypoints in the mission to clear all obstacles, even if you have to do it manually.

The lateral GPS accuracy is the real problem in tight spaces. It can also easily be off by up to 25 feet from one flight to the next, so program that 25 foot variance in, away from all obstacles, on both sides, too, at your new adjusted height.

 

[shb]

Not exactly the same place. Longitude in one is 120.58729 and 120.587298 in the other.
The abs_alt is actually the rel_alt because the only elevation DJI can measure is relative to the launch point.
For greater accuracy of the relative elevation changes measured by the barometer, let it warm up to better calibrate itself, by letting it sit for a minute or two, before arming the motors, after turning the battery on, even if the Homepoint has already been acquired. As noted by others, the barometric accuracy can easily be off by 10-25 feet, so add an additional 25 feet to all your waypoints in the mission to clear all obstacles, even if you have to do it manually.

The lateral GPS accuracy is the real problem in tight spaces. It can also easily be off by up to 25 feet from one flight to the next, so program that 25 foot variance in, away from all obstacles, on both sides, too, at your new adjusted height.

Well, if you want to goto 5 decimal points, it's not exact! Pretty close. Look at the pictures. I also noticed the lateral position looked off a little. I wonder if you go to altitude and wait a minute or two, if the position will be more accurate. So why is the abs altitude so different? Really, we don't care about that anyway, since only relative altitude is used in waypoint coordinates. But I'll try your suggestion.

 

[GadgetGuy]

Well, if you want to goto 5 decimal points, it's not exact! Pretty close. Look at the pictures. I also noticed the lateral position looked off a little. I wonder if you go to altitude and wait a minute or two, if the position will be more accurate. So why is the abs altitude so different? Really, we don't care about that anyway, since only relative altitude is used in waypoint coordinates. But I'll try your suggestion.

That’s 5 decimal points over the earth! Pictures are quite different, too, but that doesn’t explain the .SRT differences in elevation. The .SRT data also are not necessarily accurate. Drone elevation could still be the same in both flights. Barometer could just be uncalibrated in one or the other.

Just need to let the barometer calibrate itself for greater accuracy. Being at altitude won‘t help. It just wastes battery power, because the drone is also hovering with the motors armed. Also, make sure the Home Point is fully acquired with the auditory prompt before launching. Between those two issues, you’ll have the least variance in position, both laterally and vertically.

 

[Meta4]

I found out the srt file gives relative and absolute altitude. When I checked to mission, the absolute altitude was 25 feet lower today. I was shocked, and there's no way to automatically raise every waypoints elevator by 25 feet. So this is a big setback. Here's a screen shot of the exact same place on different days.

Don't pay any attention to what DJI label as Absolute Altitude.
It cannot be used for anything and as you found, it will be different on different days.

The abs_alt is actually the rel_alt because the only elevation DJI can measure is relative to the launch point.

Absolute altitude is not related to relative altitude.
Relative altitude is the only height that means anything and the only one you can use.
It's the height you see on your screen and is relative to the launch point (which = zero).

The height that DJI labels as Absolute Altitude or sometimes GPS altitude (although it doesn't come from GPS), is completely useless.
It's an approximation of the height above sea level, but because DJI used a faulty algorithm to calculate it from air pressure, it's completely wrong and the size of the error changes from day to day.

Here are examples from stills and video to demonstrate:

Both altitudes are in metres.
The relative altitude is close to correct - the drone was about 20-something metres above the launch point.
But I'm pretty sure that the absolute altitude is not correct .. the drone was not 90+ metres below sea level.

DJI attempted to calculate absolute altitude from air pressure, but didn't account for the normal variation of air pressure due to weather.

 

[Meta4]

I wonder if you go to altitude and wait a minute or two, if the position will be more accurate.

It won't make any difference that you'd notice.

So why is the abs altitude so different?

See post #16.