Mavic Air never returned

[TodayDee72]

Thank you for everyone's input. Lesson's all learned and lots of flight time needed so I don't experience the same errors with my Air2s.

 

[Rangerider]

Just thinking about one or the responses about flying downwind here. Doesn't the on board flight computer take into account the the wind for calculating the RTH time? Wouldn't the calculation include an additional flight time for return into the wind in order to make it back to the home point with a little to spare? I normally fly down to about 25% just to be easier on the battery and store the battery at 65% for more than a couple of days storage.

 

[snowghost]

Just thinking about one or the responses about flying downwind here. Doesn't the on board flight computer take into account the the wind for calculating the RTH time? Wouldn't the calculation include an additional flight time for return into the wind in order to make it back to the home point with a little to spare? I normally fly down to about 25% just to be easier on the battery and store the battery at 65% for more than a couple of days storage.

That's a good question. The flight logs as analyzed in AirData show the wind speed and direction aloft. So the bird must be aware of the wind and direction affecting RTH time.

 

[Rangerider]

Maybe you should post a reward for it? Possibly a gopher carried it away? Maybe think about putting some strobes on it for your next drone making it easier to spot especially in low light? Strobes have a long battery life and are easily visible.

 

[Rangerider]

I've often seen large birds 'eyeing' my drone thinking it might be their next meal so I tend to return quickly if a menacing bird seems interested in my drone! Around here we have crows, hawks, eagles, vultures, etc in Colorado. Maybe I should put some mtn lion scent on my drone to avoid pesky predators?

 

[Meta4]

Just thinking about one or the responses about flying downwind here. Doesn't the on board flight computer take into account the the wind for calculating the RTH time?

It factors in distance but not the wind you might encounter on the return flight.
The pilot has to do some of the thinking.
If your return will involve a hard battle against a strong headwind, you need to make a big allowance for the reduced speed.
Or revise your planned flight to avoid getting into that situation.

It always helps to start with a full battery too rather than one with less than 50% capacity like the one in this thread.

 

[Rangerider]

That's a good question. The flight logs as analyzed in AirData show the wind speed and direction aloft. So the bird must be aware of the wind and direction affecting RTH time.

That's what I would logically think but I don't know the answer. It seems to me also that even with a flight involving a partially charged battery (which I do often when not wanting to fly much) that the flight management computer would take into account all factors regarding RTH including time/distance/wind and remaining battery charge sufficient to make a successful RTH with a little to spare. However, interruption of the RTH with manual commands would subvert the completion of RTH with adequate safety factor. I've never tested that system to see how well it works so far! Sorry for your lost drone. This incident could also be a good argument for 'water-proofing' your drone but I'm not doing that for now.

 

[Rangerider]

It factors in distance but not the wind you might encounter on the return flight.
The pilot has to do some of the thinking.
If your return will involve a hard battle against a strong headwind, you need to make a big allowance for the reduced speed.
Or revise your planned flight to avoid getting into that situation.

It always helps to start with a full battery too rather than one with less than 50% capacity like the one in this thread.

Hmm. That's nice to know. I see this as a flaw in the RTH program logic if you ask me. I do think about the wind as a factor while flying but it seems to me that it would be an easy algorithm to include an average wind correction into the RTH logic to 'help' out the pilot which would calculate a direct RTH with average wind calculation to do a successful RTH with just a little extra battery life. I can see one possible reason being that if the pilot had repositioned his/her location during the flight (driving around?) that the RTH can be overridden manually but 99% of the time that would not be the case. When I get a low battery warning I return immediately and set a 'reasonable' battery warning to give me time to return. BTW, I have seen some +/- 30 ft takeoff elevation after flying around. I tend to hover a bit low to the ground after TO to let the GPS locate my RTH position/altitude.

 

[MAvic_South_Oz]

Doesn't the on board flight computer take into account the the wind for calculating the RTH time? Wouldn't the calculation include an additional flight time for return into the wind in order to make it back to the home point with a little to spare?

It factors in distance but not the wind you might encounter on the return flight.

As Meta4 mentioned, no, there are no calculations for wind in the drone 'brain'.

As well as distance from home point, there is also calculation for descent needed, or ascent if very low in a valley) to get back to the HP.

The Airdata windspeed / wind direction info is gathered from another source, and not an aircraft calculation.

I too like to RTH (manually usually) with around 20 - 25% battery remaining, so set my low battery warning to 30% and critical warning to 20%.
Can't remember getting a critical warning in a loooonnng time, that is more for me to start looking for a place to land if required . . . generally not an issue as most flights are within 600m, 700m or closer.

 

[Rangerider]

Just thought of another factor when returning into a strong headwind while running low on battery. The old adage for flying is go slow with a tailwind and fast into a headwind which reduces your flight time as much as possible even though you get a bigger battery drain at high power. think of it this way, if you fly at reduced speed you more nearly match the windspeed and have a longer flight time which is not good. Not sure what the best speed is into the wind but I have noticed that a speed of about 15 mph seems to give the longest flight time all things considered. Probably the most efficient speed for the MP2? I know this doesn't help the lost drone owner but maybe will in the future?

 

[Rangerider]

As Meta4 mentioned, no, there are no calculations for wind in the drone 'brain'.

As well as distance from home point, there is also calculation for descent needed, or ascent if very low in a valley) to get back to the HP.

The Airdata windspeed / wind direction info is gathered from another source, and not an aircraft calculation.

I too like to RTH (manually usually) with around 20 - 25% battery remaining, so set my low battery warning to 30% and critical warning to 20%.
Can't remember getting a critical warning in a loooonnng time, that is more for me to start looking for a place to land if required . . . generally not an issue as most flights are within 600m, 700m or closer.

Yeah, better to do a controlled landing before the battery goes dead than try to keep on going and try gliding it in! Using the camera for a suitable landing spot beats the dead stick scenario. I haven't tried that either yet. I don't know the auto-rotation speed for the drone but that's moot with a dead battery anyway.

 

[MAvic_South_Oz]

Yeah, better to do a controlled landing before the battery goes dead than try to keep on going and try gliding it in! Using the camera for a suitable landing spot beats the dead stick scenario. I haven't tried that either yet. I don't know the auto-rotation speed for the drone but that's moot with a dead battery anyway.

True, the only worry with manual landing in such a case is video transmission might fail, then drone - controller signal might be lost, and RTH failsafe kicks in.
Probably wouldn't happen if there was signal LOS to the drone, but if it gets too low behind terrain or a lot of vegetation, it would likely happen.

It might be best to practice long flights by setting failsafe to land, at least if you are descending onto say a grassed or open area, it should lose contact but still land.
Alternatively set to hover, you can either get there fast to regain signal if possible, or it will reach battery auto land level.
Providing wind isn't an issue (blow-away) then that should be the safest option.

Well, safest really is flying VLOS, or very close to it.
I honestly can't ever recall losing signal, a couple of times little video flickers, but that's it.

 

[Meta4]

Not sure what the best speed is into the wind but I have noticed that a speed of about 15 mph seems to give the longest flight time all things considered. Probably the most efficient speed for the MP2?

Do you want the longest flight time or distance?
Longest flight time with your MP2 is achieved at a steady 7 metres/sec (15.5 mph) in still air.
But in a decent headwind that might never get you home.
Max distance is achieved at close enough to top speed in P-GPS mode with obstacle avoidance disabled.

 

[Meta4]

I see this as a flaw in the RTH program logic if you ask me. I do think about the wind as a factor while flying but it seems to me that it would be an easy algorithm to include an average wind correction into the RTH logic to 'help' out the pilot

Are you assuming that people always make straight line flights out and back?
How would you calculate for winds that are at an angle to the flight path?
Here's a flight I did last week that involved significant winds.

How simple would it be to program the drone to keep track of the wind data for a flight like this and calculate how the return would be affected by wind?
It's a lot easier for it to calculate RTH battery requirements based on distance (and probably a small safety factor) and leave the flyer to do the extra calculating for anything more than a simple still air flight.

BTW, I have seen some +/- 30 ft takeoff elevation after flying around. I tend to hover a bit low to the ground after TO to let the GPS locate my RTH position/altitude.

If you are saying that you've observed the indicated height on screen drift over the time of a flight, that's common and normal with a barometric sensor used to measure height.
GPS is not used to establish altitude, and hovering won't make any difference.

 

[Faster]

Are you assuming that people always make straight line flights out and back?
How would you calculate for winds that are at an angle to the flight path?
Here's a flight I did last week that involved significant winds.

How simple would it be to program the drone to keep track of the wind data for a flight like this and calculate how the return would be affected by wind?
It's a lot easier for it to calculate RTH battery requirements based on distance (and probably a small safety factor) and leave the flyer to do the extra calculating for anything more than a simple still air flight.


If you are saying that you've observed the indicated height on screen drift over the time of a flight, that's common and normal with a barometric sensor used to measure height.
GPS is not used to establish altitude, and hovering won't make any difference.

Sorry for the stupid question. If the drone would be able to track the wind data at all times and compare them with its heading position using the GPS. When it flies to any direction couldn't be able to calculate how the winds would affect RTH? It knows beforehand where the home point is. Thanks.

 

[sar104]

Sorry for the stupid question. If the drone would be able to track the wind data at all times and compare them with its heading position using the GPS. When it flies to any direction couldn't be able to calculate how the winds would affect RTH? It knows beforehand where the home point is. Thanks.

The FC in all the current DJI models does compute wind velocity (speed and direction), and in a fairly constant wind field it would be simple enough to factor that into the smart RTH calculation. However, wind velocity is generally a non-linear function of height, and often varies rapidly with time at any given height. As a result the success of that strategy would be limited, which is probably why it isn't implemented.

 

[Faster]

The FC in all the current DJI models does compute wind velocity (speed and direction), and in a fairly constant wind field it would be simple enough to factor that into the smart RTH calculation. However, wind velocity is generally a non-linear function of height, and often varies rapidly with time at any given height. As a result the success of that strategy would be limited, which is probably why it isn't implemented.

I never thought about the wind's speed and direction variability. Thanks for the explanation.

 

[Kayak Rob]

How simple would it be to program the drone to keep track of the wind data for a flight like this and calculate how the return would be affected by wind?
It's a lot easier for it to calculate RTH battery requirements based on distance (and probably a small safety factor) and leave the flyer to do the extra calculating for anything more than a simple still air flight.

Wattage currently being used and a display of estimated battery percent, or distance to home battery percent.
Not that DJI would even implement such a feature.

Let me explain:
I have a trolling motor for my kayak. Inside the lithium battery is also a GPS.
My trolling motor console (controller) display shows me wattage, speed, percent of battery,
and a calculation of miles I have left at the speed and wattage I am using.
For example... at 3mph no wind and current it may show 40 miles remaining.
............................at 5mph against wind and current it may show 8 miles remaining.
And thusly... if my battery drains down to 50% later in the day... the above numbers drop accordingly.

So.... if the drone had a way to measure wattage being used by the motors, it could calculate percent of battery being used and relative battery remaining.
IE:
Basically....
Battery says it is at 60%
But right now, with strong head winds (higher wattage)... your battery has 30% remaining (at current wattage drain).

Better yet... a calculation of drone's current position, RTH position, wattage/battery drain...
and a representation of battery percentage remaining when landed at "home."

My trolling motor... when it gets down to 30% battery....
Displays.... "Drive Slow."

 

[Skwirl]

Shouldn't it read start paddling, lol.

 

[slup]

So.... if the drone had a way to measure wattage being used by the motors, it could calculate percent of battery being used and relative battery remaining.

Better yet... a calculation of drone's current position, RTH position, wattage/battery drain...
and a representation of battery percentage remaining when landed at "home."

As already said ... not much is constant for very long when flying drones, measuring like you propose will easily result in information that constantly change drastically over time.