Launch and return to boat

[RTKD]

So just for an example of what NOT to do take a look at this:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

The biggest mistake here was trying to bring the drone in facing you. The drone should be pointed away from you, that way your right and left aren't backwards. She got her right and left switched, and ended up in the briny deep.

You also want to catch and release from the stern, with the boat facing up wind.

 

[bsartist]

If I may, I'd like to add some compass terminology specifics to maybe clear up some confusion and hopefully explain why @Meta4 says some of the things he says.

Compass variation is the difference between true north and magnetic north at your particular location on the earth. While it does drift over the years, it's a pretty much a fixed number based on where you are.

Compass deviation is how much the compass is off (deviates from magnetic north) due to the presence of ferrous metals nearby.

When people talk about calibrating when moving so many miles, they are likely thinking that you are correcting for variation. I have not seen this specially documented but my strong bet is that these drones have a variation table coded into them and therefore know the variation of any location you go to. Hence calibration is not addressing variation and therefore unnecessary when changing location.

That leaves deviation. When calibrating a compass for deviation, you are compensating for ferrous metals around the compass that are in a fixed location relative to the compass. Think of the compass at the helm of a boat. You've got ferrous metal in stuff fixed to your dash around the compass which can deflect the needle away from magnetic north. A good old fashioned magnetic compass actually has little bits of adjustable metal that you position with screws to counteract the metal on your dash and return the needle to pointing at magnetic north. Once you set them, you leave them alone unless you add something ferrous permanently to your dash. And you avoid laying anything metallic nearby temporarily as it adds deviation back in. You wouldn't want to recalibrate based on your iPhone laying there only to have to recalibrate back when you pick it up. Or worse: recalibrate for the iPhone and then take the iPhone away and leave the old calibration. If you do that your compass is off.

So my other strong bet is that compass calibration on these drones is meant to correct for deviation caused by ferrous material inside the drone only. That stuff shouldn't change and that's why you don't need to calibrate the compass all the time and especially why you should never calibrate your compass in the presence of other ferrous metals that won't be there after you take off. You have deliberately added an erroneous deviation compensation. Once airborne and that temporary source of deviation is removed, your compass is off.

Now, I'm not sure why it requires calibration every so often. It's probably an electronic flux compass of some kind which maybe can get thrown off. Maybe if it ends up in a magnetic field (not the same as near ferrous but non magnetized metal).

So if my strong bets are correct, that's why calibrating based on geo location is unneeded while calibrating in the close proximity of ferrous metals is detrimental and a bad idea.

Bill

 

[DanMan32]

Not that this is a thing... but how would this affect anything?

Because VPS affects position. It normally would be more accurate than GPS and so would be more likely used when in range.
If the reference is believed to be stationary but is in fact moving, then the device would move with the reference.

 

[tcope]

Because VPS affects position. It normally would be more accurate than GPS and so would be more likely used when in range.
If the reference is believed to be stationary but is in fact moving, then the device would move with the reference.

VPS does not position the Mavic when the Mavic has a good GPS lock and then it only works when landing.

 

[Meta4]

The boat flights are further complicated by potential IMU error problems related to various metal components or accessories on the boat.

The IMU is not affected by "metal".
You are confusing that with the compass which is affected by being placed close to steel.

 

[Meta4]

Because VPS affects position. It normally would be more accurate than GPS and so would be more likely used when in range.
If the reference is believed to be stationary but is in fact moving, then the device would move with the reference.

For VPS to work, it has to have a surface with a recognisable pattern or texture.
It doesn't have that with a water surface.
VPS is just going to give up.

 

[dbyantis]

Hi I'm planning an outing on a friend's 50 foot aluminium boat - plenty of steel as well no doubt. I expect I'll be able to lock into ~10 - 15 satellites but not sure about compass calibration. If I have clear viability; no wind etc am I going to be able to operate without risk of rogue flyaway etc - I'm not for example wether the controller will allow me to fly if steel upsets the compass? Thanks

Check you RTH settings, you do not want it to fly back to the launch point in a moving boat, In the event of a RC disconnect, you probably want to hover so you can drive the boat to retrieve. Or launch from land, get in the boat and have fun.

 

[SteveCA]

I’ve flown quite a bit from boats and have observed the following:

I almost always get a compass calibration warning but it goes away as soon as the drone lifts off. The rest of the flight is just fine.

I always hand catch after turning off vps. It works fine but wear gloves and practice on land until it is a no brained.

After take off, make sure you hover long enough to set the home point in a place where you will be able to see the drone even if your boat has moved.

If you can’t stop the boat, it may be helpful to have someone else’s hand catch the drone while you match the boat speed and direction. Obviously you should be very adept at flying before you try this.

 

[DanMan32]

For VPS to work, it has to have a surface with a recognisable pattern or texture.
It doesn't have that with a water surface.
VPS is just going to give up.

It does when you have ripples/waves on the water surface.
Hence the moving ripples/waves deceiving the AC into moving when it thinks it is stationary based on the water surface pattern.

 

[DanMan32]

VPS does not position the Mavic when the Mavic has a good GPS lock and then it only works when landing.

Are you sure about that?

 

[Meta4]

It does when you have ripples/waves on the water surface.
Hence the moving ripples/waves deceiving the AC into moving when it thinks it is stationary based on the water surface pattern.

Give up on this one.
The VPS just can't lock on to any recognisable features on the water's surface (because there aren't any).
The Phantom knows if it's moving because it's gyro sensors provide that data - not the VPS which is only used for positioning when conditions are suitable.

 

[tcope]

Are you sure about that?

Yes.

 

[SteveCA]

Yes.

I have a friend that found that to be false. He was flying his P4 low over a stream with moving water. with no stick input, while hovering, his drone drifted.... sideways right into an obstacle and drowned. After analysis, it was determined that the vps sensors fooled the drone into thinking it was moving when it was hovering.

The rest is history for that P4. Expensive lesson. Even DJI suggest turning off vps when flying low over water

 

[tcope]

with no stick input, while hovering, his drone drifted.... sideways right into an obstacle and drowned. After analysis, it was determined that the vps sensors fooled the drone into thinking it was moving when it was hovering.

The information I posted was correct but to what you stated... with no stick movement the drone would hover but instead it moved into a tree. But you then state that that the sensors tricked it into thinking it was moving. If tricked into thinking it was moving... it was not moving. So it would not move into the tree.

 

[ColinF]

Do not fly with a compass problem. Learn how to hand launch and land. This can be tricky if the boat is rocking at all. Wear gloves.

I second the wear gloves... scraggly dried out old hand aside, the purple still healing scar is from a Mavic 2 prop. On the plus side, it cuts like a deli slicer, perhaps a secondary use case for the drone.

 

[DanMan32]

All I know is on a windy day over water, my M2 was drifting like crazy which had me worried. It was rather low altitude, thought I admit I'm not quite sure if it was in VPS range. The drift correlated with water currents and waves related to wind. I brought it home which beelined back to me.
So based on that, I can only conclude the M2 was tracking the moving water currents, assuming they were stationary patterns, and occasionally correcting it's drift based on other inputs besides VPS.

 

[Meta4]

All I know is on a windy day over water, my M2 was drifting like crazy which had me worried. It was rather low altitude, thought I admit I'm not quite sure if it was in VPS range.
... So based on that, I can only conclude the M2 was tracking the moving water currents, assuming they were stationary patterns, and occasionally correcting it's drift based on other inputs besides VPS.

Based on what?? A guess and no evidence?
If you don't even know if it was in VPS range, why are you trying to claim it was VPS?
It's not hard to tell.
Post your flight data and it will be immediately obvious if VPS was causing issues.

Go to DJI Flight Log Viewer - Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
Come back and post a link to the report it gives you.

 

[IronIce]

Error = no flying