Is it possible to fly a drone from the deck of a steel ship or boat?

[CactusJackSlade]

My brother does this all the time working for a few conservation groups that use large ships. Calibrate compass on land, hand launch and catch is what he has said to me.

 

[dave canada]

sparkie will do it easy mavic pro noway

 

[Meta4]

I would not recommend it. I knew a guy that flew his Inspire near a 250' steel water tower and it glitched out and crashed on top of the water tower.

Unless you mean closer than a few feet, flying near a large steel structure of ship won't have any effect on your flying.
Flying behind a large steel structure is a different matter, that will block your signal and initiate RTH.

if the vessel has operating radar (and most are required to do so while underway - in must cases multiple redundant open array radars) it is a strong signal that although a different bandwidth/spectrum from your control channels will likely cause extreme interference (blanking or similar intermittent signal loss when the transmitter sweeps your direction). same with any radio antenna masts like AIS that continually transmits data as to the vessels position. these interference sources to me are more concerning than the fact that the vessel may have ferrous structural components.

I fly a couple of hundred ships each year and have never noticed any issue from radar, radios etc.
Here's one of a few that I flew yesterday.

 

[OurAngryBadger]

Unless you mean closer than a few feet, flying near a large steel structure of ship won't have any effect on your flying.
Flying behind a large steel structure is a different matter, that will block your signal and initiate RTH.

I fly a couple of hundred ships each year and have never noticed any issue from radar, radios etc.
Here's one of a few that I flew yesterday.

He had LOS to the Inspire, didn't fly it behind the tower. It was about 15 ft in front of the tower and maybe 10 ft above it, when he tried to get a little closer to the tower he lost signal and the Inspire drifted down and into the top of the tower.

 

[Meta4]

He had LOS to the Inspire, didn't fly it behind the tower. It was about 15 ft in front of the tower and maybe 10 ft above it, when he tried to get a little closer to the tower he lost signal and the Inspire drifted down and into the top of the tower.

That sounds a little odd.
If you fly so close that the magnetic field of the steel has any effect, the effect is on the compass.
It doesn't cause loss of signal or losing height.
Without the data, it's hard to know what really happened.

 

[WilliamVargas]

I have repeatedly launched a drone from different ships. Calibrate the compass on land, launch with your hand and everything will be great.

 

[AirShark]

I fly off of an aluminum boat and haven't any real issues other than my nerves. I definitely recommend hand catching and make sure you set your RTH position to controller and elevation above the tallest point on the ship. Make sure you return with enough battery remaining in case you need to deal with winds or a drifting boat. I also fly with a "Getter Back" strapped to a leg just in case. I don't frequently sail in water deeper than 100'.

 

[skiptv]

If you want your drone to fly way, then take off from a metal ship. A metal ship or metal on the ground will cause compass error, if you take off with a compass error your drone does not really know where its going, thus a fly away....

 

[BJR981S]

Does anyone successfully fly from the deck of a steel ship or boat?

Or does the steel just send the compass crazy?

Surely if you calibrated at deck level then once you took off away from the magnetic field generated by the ship it would cause major issues?

I know that they perform drone inspections of oil rigs and platforms, which must be similar but not sure how they get around the problem?

Its not a problem as long as you don't ask the drone to fly to a Waypoint or RTH. The compass may very well be out of wack but normal flight is controlled by the 6G Gyro and Accelerometers.

Of course if you lose connectivity its likely to fly away.

Just fly it like you don't have GPS and fly in Atti Mode.

 

[sar104]

Its not a problem as long as you don't ask the drone to fly to a Waypoint or RTH. The compass may very well be out of wack but normal flight is controlled by the 6G Gyro and Accelerometers.

Of course if you lose connectivity its likely to fly away.

Just fly it like you don't have GPS and fly in Atti Mode.

That's not correct. The rate gyros are the primary source of yaw data in flight, but the absolute value is determined by the compass. The IMU yaw value is initialized by the compass, and if that is incorrect at power up then the FC will be working with an incorrect yaw value. That's the source of almost all compass/yaw errors and resulting uncontrolled flight.

 

[OldGoatMTB]

I wonder what happens when it's time to land & you can't avoid some other magnetic area that upsets the compass ...

Dunno, it I landed on the roof of my van for fun once, and then it refused to re-arm and take off.

 

[Gringorio]

If you want your drone to fly way, then take off from a metal ship. A metal ship or metal on the ground will cause compass error, if you take off with a compass error your drone does not really know where its going, thus a fly away....

No, not really. There may be a compass error message at take off, but resolved while a hover away from the ship after takeoff. I've done hundreds of flights, many with this message, but with all successful flights.

 

[slup]

No, not really. There may be a compass error message at take off, but resolved while a hover away from the ship after takeoff. I've done hundreds of flights, many with this message, but with all successful flights.

That conviction should be handled with care ... if the compass is disturbed during startup when the IMU initialize Yaw & magYaw still will agree within the magnetic interfered area. This can be seen in your GO4 or Fly app as the AC heading direction on the map doesn't match the reality. But once airborne & away from the magnetic interfered area Yaw & magYaw will disagree & you will face the consequences of the toilet bowl effect.

 

[Gringorio]

That conviction should be handled with care ... if the compass is disturbed during startup when the IMU initialize Yaw & magYaw still will agree within the magnetic interfered area. This can be seen in your GO4 or Fly app as the AC heading direction on the map doesn't match the reality. But once airborne & away from the magnetic interfered area Yaw & magYaw will disagree & you will face the consequences of the toilet bowl effect.

I agree, but have not yet experienced this even after hundreds of flights. Though I have to say the Mavic Pro 2 seems to be more reliable than the Phantom 4 Pro. Not sure why.

 

[slup]

I agree, but have not yet experienced this even after hundreds of flights. Though I have to say the Mavic Pro 2 seems to be more reliable than the Phantom 4 Pro. Not sure why.

When it comes to the M2P/Z, think they have the ability to correct & equal out a disagreement between Yaw & magYaw when it occures by aligning Yaw to magYaw if I'm remembering correctly ... heard it from @sar104 I think ... perhaps he can correct me or explain more.

 

[sar104]

When it comes to the M2P/Z, think they have the ability to correct & equal out a disagreement between Yaw & magYaw when it occures by aligning Yaw to magYaw if I'm remembering correctly ... heard it from @sar104 I think ... perhaps he can correct me or explain more.

Yes - the Mavic 2 detects, and corrects after takeoff, an incorrectly initialized IMU yaw:

Compass error demonstration

Ok my apologies, I don’t own a Mavic but I know how my phantom standard and i1 reacted in these exact same locations, pix below. With a compass error. I was totally wrong for flying in these areas but I clearly remember how my drones reacted after take off. I had to walk out of the block to keep...

mavicpilots.com

 

[BJR981S]

That's not correct. The rate gyros are the primary source of yaw data in flight, but the absolute value is determined by the compass. The IMU yaw value is initialized by the compass, and if that is incorrect at power up then the FC will be working with an incorrect yaw value. That's the source of almost all compass/yaw errors and resulting uncontrolled flight.

So you say.

The Yaw Accelerometers and Gyro are 1,000% more accurate than the compass. The compass is used primarily for reporting position back to the App so you have a directional indicator and for calculating the course for Waypoint missions and RTH. The IMU, with the 3 axis Gyro and the 3 Axis Accelerometers are calibrated at startup. There is no mechanism to change one axis on the Gyro. It would be totally pointless and introduce large errors.

The Gyro and Accelerometer have no bearing on the compass vector.

There are many manufacturers that produce non Compass non GPS drones and they fly just fine and are as sensitive to yaw as DJI.

Check out some 3D heli Videos on youtube. They use 6 axis IMUs and no compasses.

I have several DJI flight controller Helicopters and Drones. Naza, Wookong and A2-3.1


Cheers

 

[BJR981S]

Yes - the Mavic 2 detects, and corrects after takeoff, an incorrectly initialized IMU yaw:

Compass error demonstration

Ok my apologies, I don’t own a Mavic but I know how my phantom standard and i1 reacted in these exact same locations, pix below. With a compass error. I was totally wrong for flying in these areas but I clearly remember how my drones reacted after take off. I had to walk out of the block to keep...

mavicpilots.com

My goodness, The is no relative position for the Gyro and Accelerometer. They report on change in attitude. The IMU is calibrated via a manual process. It is reset at power-up . You can't change it or calibrate it or anything else in flight. It is what it is.

Think about what you just said. You ave a device that you say is incorrectly initialised. That would mean that all 3 axis are incorrectly initialised. In flight it would be moving in all 3 axis. There is no "Witness Mark" you can use to centre or correct anything, in any axis.

 

[sar104]

So you say.

The Yaw Accelerometers and Gyro are 1,000% more accurate than the compass. The compass is used primarily for reporting position back to the App so you have a directional indicator and for calculating the course for Waypoint missions and RTH. The IMU, with the 3 axis Gyro and the 3 Axis Accelerometers are calibrated at startup. There is no mechanism to change one axis on the Gyro. It would be totally pointless and introduce large errors.

The Gyro and Accelerometer have no bearing on the compass vector.

There are many manufacturers that produce non Compass non GPS drones and they fly just fine and are as sensitive to yaw as DJI.

Check out some 3D heli Videos on youtube. They use 6 axis IMUs and no compasses.

I have several DJI flight controller Helicopters and Drones. Naza, Wookong and A2-3.1


Cheers

My goodness, The is no relative position for the Gyro and Accelerometer. They report on change in attitude. The IMU is calibrated via a manual process. It is reset at power-up . You can't change it or calibrate it or anything else in flight. It is what it is.

Think about what you just said. You ave a device that you say is incorrectly initialised. That would mean that all 3 axis are incorrectly initialised. In flight it would be moving in all 3 axis. There is no "Witness Mark" you can use to centre or correct anything, in any axis.

Every year or so it seems that you come back and post this stuff again, and then go quiet. This year you are still wrong. What baffles me is that, quite apart from the fact that these are all established technologies that you can read about in the literature, the flight log data show all these things happening. You can see the IMU yaw being initialized by the magnetometer data, and then being corrected by the local declination when the FC gets an GNSS position lock. This is not hypothesizing.

As for the rest of your posts, I really don't even know where to start because you are just massacring basic physics and maths. You don't even understand what is meant by "initialize", apparently. It has nothing to do with calibration. And no - all three gyro axes are not the same - two of them are fixed by an absolute measurement that the IMU can make. Initial pitch and roll are measured directly by the accelerometers. Only the yaw axis is indeterminate on startup. And so, just with any gyro compass, that one has to be initialized by a compass reading.

 

[BudWalker]

Every year or so it seems that you come back and post this stuff again, and then go quiet. This year you are still wrong. What baffles me is that, quite apart from the fact that these are all established technologies that you can read about in the literature, the flight log data show all these things happening. You can see the IMU yaw being initialized by the magnetometer data, and then being corrected by the local declination when the FC gets an GNSS position lock. This is not hypothesizing.

As for the rest of your posts, I really don't even know where to start because you are just massacring basic physics and maths. You don't even understand what is meant by "initialize", apparently. It has nothing to do with calibration. And no - all three gyro axes are not the same - two of them are fixed by an absolute measurement that the IMU can make. Initial pitch and roll are measured directly by the accelerometers. Only the yaw axis is indeterminate on startup. And so, just with any gyro compass, that one has to be initialized by a compass reading.

I bet the next post contains the rant about confirmation bias.