DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Over water risk reduction

As far as I know, as long as the drone isn't actually hovering, it won't descend, unless by malfunction or hitting something.

Am I wrong?
 
Water is reflective in the visible spectrum but the drone has an infrared sensor on the bottom of it. Water becomes increasingly opaque due to light absorption in the infrared spectrum (i.e. not reflective or transparent).
The-spectrum-of-absorption-coefficient-of-liquid-water-2.png
 
Water is reflective in the visible spectrum but the drone has an infrared sensor on the bottom of it. Water becomes increasingly opaque due to light absorption in the infrared spectrum (i.e. not reflective or transparent).
The-spectrum-of-absorption-coefficient-of-liquid-water-2.png
Water absorbs IR light quite well, but it still manages to reflect it a reasonable amount. Since most IR that penetrates through the surface is absorbed what gets reflected from the surface is much more noticable.

 
Water absorbs IR light quite well, but it still manages to reflect it a reasonable amount. Since most IR that penetrates through the surface is absorbed what gets reflected from the surface is much more noticable.

You would have to know the frequency the infrared sensor works in to know for sure what it would see. I don't know if it is near-infrared, middle, or whatever. I've flown my drone over lakes just a couple of feet above the water with no problem but the lake isn't exactly clear. I would probably be more careful if it was over a swimming pool or something like that.
 
You would have to know the frequency the infrared sensor works in to know for sure what it would see. I don't know if it is near-infrared, middle, or whatever. I've flown my drone over lakes just a couple of feet above the water with no problem but the lake isn't exactly clear. I would probably be more careful if it was over a swimming pool or something like that.
Yes, absorption is quite frequency dependant. I think the sensors use NIR as I can see the emitter with my eyes, and a typical phone camera IR filter doesn't block much of the emitter light at all. My MM has detected the river bed in a shallow stream (about 20cm deep and very clear water) so it was getting enough transmitted light through effectively 40cm of water to function.
 
You would have to know the frequency the infrared sensor works in to know for sure what it would see. I don't know if it is near-infrared, middle, or whatever. I've flown my drone over lakes just a couple of feet above the water with no problem but the lake isn't exactly clear. I would probably be more careful if it was over a swimming pool or something like that.

Certainly it's NIR. Wavelengths longer than 1000nm would require InGaAs photodiodes, not silicon, and InGaAs is quite a bit more expensive.

Small pcb-mount proximity sensors that I've used recently have 850nm VCSEL lasers internally. These proximity sensors typically do a time-of-flight measurement. There is a "ping" of light, and the first return is the one that is used. So if there was a return from both the top of the water and the river bottom, the top of the water would be registered as the distance every time.
 
Water is reflective in the visible spectrum but the drone has an infrared sensor on the bottom of it. Water becomes increasingly opaque due to light absorption in the infrared spectrum (i.e. not reflective or transparent).
The main point is not whether the IR sensor can detect the water surface. Base on experience, the sensor can do so in most cases. In those relatively rare cases in which the water surface cannot be detected, the drone will still be able to hold the altitude base on the barometer readings. It may drift up and down a bit because of the limited precision of barometer but it will NEVER go into the water by itself as rumoured. If it does, there must be some other reasons. A common cause I have seen in this forum is the pilot holding the throttle stick down.
 
This is a test I have done earlier with my M2P under controlled environment - a bucket filled with water. The VPS figure at the bottom of the screen of the GO APP is the distance from the water surface detected by the IR sensor at the belly of the craft

The result shows that the sensor is able to detect the water surface quite reliably even if the water is super clear and the surface is dead still. When the surface is choppy, I am yet to see a miss.

 
Last edited:
  • Like
Reactions: passedpawn
The lowest I flew over water was below 1 meter (original Mavic Pro). I started on land to get the drone as low as possible without landing, then started flying. The drone would not be recoverable when it would have crashed in the water, so that was the risk I accepted...
 
On your video clip of the ocean, if you used a filter, what was it? Liked the deep blue waters and sky line.
I've never used a filter, I think they're a waste of money and that they constrict your camera. You can do all that in editing altho I rarely do and did not touch this video in anyway. This is raw from the SD card
 
Reading the manual I get the impression that the infrared system is used during "regular flight" only when there's no GPS-signal available (for example when flying indoors), and during emergency landing when the drone is checking the ground if it looks flat enough. Isn't that right?
That would mean that hovering and flying over water, snow, ice or whatever is not a problem as long as you have got a GPS-signal that is strong enough. Correct?
 
Reading the manual I get the impression that the infrared system is used during "regular flight" only when there's no GPS-signal available (for example when flying indoors), and during emergency landing when the drone is checking the ground if it looks flat enough. Isn't that right?
That would mean that hovering and flying over water, snow, ice or whatever is not a problem as long as you have got a GPS-signal that is strong enough. Correct?
Not really,
GPS is never used for altitude.
Altitude is fused from imu, baro and ultrasonic, and it's that fusing that runs into problems, specially over water. You never knows what it's up to :)

Ultrasonic is also used to avoid slamming into ground. It's initiates landing mode when below 1.2m.

I don't know how uneven landing surface is detected though. You might be correct in this one.

Starting from air 2 the vision sensors detects water, and make all sort of strange things. Sometimes it just stops if you fly to low over water. I think it is this problem TS has, but I don't know for sure.
 
  • Like
Reactions: FlyBackOU812
With the right equipment you can land on the water and take off.
IMG_0030.JPG
 
Haha,
I'm probably not the first to ask, how much wind can you fly with these floaters? (Mavic 2)
What is the max speed?
I'm doing a lot of follow me missions in high wind, and I been thinking of getting one.
The people selling them never give any details of how much the aerodynamic performance is affected.
I wonder why?
 
  • Like
Reactions: Ex Coelis
The people selling them never give any details of how much the aerodynamic performance is affected.
I wonder why?
The reason why is because you do not watch the videos of the performance and you have not tested it yourself.
And for some reason you are against the whole idea.
But that is your choice.
Mine is not taking a chance.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Members online

Forum statistics

Threads
130,978
Messages
1,558,525
Members
159,966
Latest member
rapidair