DJI Go 4 app RTH computations?

[jjckflsh]

Hi,
I have questions regarding the dynamic RTHI - Return to Home Charge Indicator on the Battery indicator in the DJI GO 4 app. How is it calculated?
Specifically, What airspeed, Climb VS, and Descent VS does it use in it's calculations? I don't believe it uses any kind of wind correction, am I correct? Also, when you push the RTH button, does it simulate a loss signal to the drone to trigger the RTH programing?
Thanks,

Jack

 

[old man mavic]

it is using GPS and compass to calculate distance from home point and then calculates how much battery it needs to get back, as you fly you will see the green spot on the battery bar move as the battery runs down in relation to the home point,it does not know about wind speed or direction, which is why people get into trouble if they have flown with the wind on the outward leg of a journey ,and then are unable to get back in RTH mode against the wind ,if this happens then there is a chance that the drone will go into low battery mode and land before it reaches home ,what you can do is as soon as you are able, is cancel RTH lower your altitude where the wind is not as strong and fly home yourself

 

[DanMan32]

It doesn't simulate signal loss.

 

[Dave Maine]

It also doesn’t calculate a return to home flight against the wind. That is a difficult calculation to do accurately, and the wind is rarely constant, either in velocity or direction.

 

[brett8883]

Hi,
I have questions regarding the dynamic RTHI - Return to Home Charge Indicator on the Battery indicator in the DJI GO 4 app. How is it calculated?
Specifically, What airspeed, Climb VS, and Descent VS does it use in it's calculations? I don't believe it uses any kind of wind correction, am I correct? Also, when you push the RTH button, does it simulate a loss signal to the drone to trigger the RTH programing?
Thanks,

Jack

We don’t totally understand that yet, however, high level: it knows the RTH horizontal speed is constant (10 m/s) and vertical speed in low battery RTH is constant (+/-3m/s). Wind may play a factor into it as well but we don’t really know. Wind doesn’t affect the time to get home and they leave a lot of room for error so it might not figure in

Low level: we actually have some in depth discussions and tests about it in one of the software mod threads

https://mavicpilots.com/threads/dji-super-patcher-for-unlocking-height-limits-nfzs-enable-galileo-satellites-more-free.68436/post-835956

 

[Dave Maine]

I think that your 10 m/s constant is true for velocity through the air. Velocity over ground, which is different, determines how quickly you get back home.
An extreme example is if the drone is returning home at 10m/s facing directly into a 10 m/s headwind, it’s speed over ground is 0, and it will never come home.
This has been empirically proved numerous times in anecdotes on this forum, usually discribed as flyaways,

 

[brett8883]

I think that your 10 m/s constant is true for velocity through the air. Velocity over ground, which is different, determines how quickly you get back home.
An extreme example is if the drone is returning home at 10m/s facing directly into a 10 m/s headwind, it’s speed over ground is 0, and it will never come home.
This has been empirically proved numerous times in anecdotes on this forum, usually discribed as flyaways,

I never said airspeed. The Mavic does not have an air speed sensor so it can’t determine airspeed directly. It only determine velocity relative to the ground. That 10 m/s is ground speed.

 

[Dave Maine]

I guess we disagree about what the DJI speed spec is measured against. All the flight controls work against the real air movements, particularly pitch to generate forward movement. That pitch is calculated not from ground speed but air speed. Calculating it from ground speed won’t work.

 

[DanMan32]

As long as it has GPS, and it needs GPS to know how to get home, it can know ground speed.

 

[Dave Maine]

Yes, but the flight control system which does the integration with gps and inertial systems, has a hard pitch limit both in p mode and sport. It must do this to avoid excessive pitch. The pitch limit is required to avoid draining the batteries too rapidly, and in the extreme leaving the drone with too little vertical force to stay aloft. Do a vector diagram for pitch and horizontal and vertical forces and you will see what is going on. All control systems in airplanes, helicopters, and boats for that matter all directly interact with the moving fluid that they are in. Movement against a fixed point is always relative to the motion of the fluid they are in.

a related example Is an aircraft stall. That occurs only when the air speed falls below a certain value. There is no direct relationship to grownspeed.

 

[Dave Maine]

The ground speed is known, calculated by the gps co-ordinates. The actual ground speed achievable is limited by the maximum pitch of the aircraft, which is determined by it’s airspeed, not ground speed.

 

[brett8883]

Yes, but the flight control system which does the integration with gps and inertial systems, has a hard pitch limit both in p mode and sport. It must do this to avoid excessive pitch. The pitch limit is required to avoid draining the batteries too rapidly, and in the extreme leaving the drone with too little vertical force to stay aloft. Do a vector diagram for pitch and horizontal and vertical forces and you will see what is going on. All control systems in airplanes, helicopters, and boats for that matter all directly interact with the moving fluid that they are in. Movement against a fixed point is always relative to the motion of the fluid they are in.

a related example Is an aircraft stall. That occurs only when the air speed falls below a certain value. There is no direct relationship to grownspeed.

You are right but RTH is different. It uses a ground speed velocity instead of pitch angle. Try it, no matter the wind speed it will always RTH at 10 m/s ground speed unless wind exceeds 10 m/s. That’s why it says not to fly in winds exceeding 10 m/s and why you’ll get wind warnings to land immediately if you do

 

[jjckflsh]

We don’t totally understand that yet, however, high level: it knows the RTH horizontal speed is constant (10 m/s) and vertical speed in low battery RTH is constant (+/-3m/s). Wind may play a factor into it as well but we don’t really know. Wind doesn’t affect the time to get home and they leave a lot of room for error so it might not figure in

Low level: we actually have some in depth discussions and tests about it in one of the software mod threads

https://mavicpilots.com/threads/dji-super-patcher-for-unlocking-height-limits-nfzs-enable-galileo-satellites-more-free.68436/post-835956

Thank you all for your replies. I appreciate your help. It would be helpful if the DJI Test engineers would weigh in to as how they tested RTH and came up with these numbers. I am not sure if we all agree that the RTH programing does not take the wind into consideration. For initial testing, I am going to say no. Taking that into consideration, the battery requirements to maintain 10 m/s would be in still air. This is for the computational display for the RTHI on the battery scale.

When you actually activate the RTH program the ball game may be different. It may try to maintain 10 m/s by decreasing pitch and increasing rotational power which would put a lot of inaccuracies into the RTHI. This is why I am asking these questions. This is a good discussion since this is the go to programing for problems.

 

[old man mavic]

all the tests and settings tend to be advisory and are in perfect ,not real world conditions ,there are too many variables, it is best to view the data shown in the app as approximate, and air on the side of caution,when doing so,my first low battery warning is set at 40% and second one at 25% ,as stated in other threads Lipos tend to drop in voltage quicker as they get lower ,and that is why you will find they recover somewhat after they have cooled and rested after use

 

[brett8883]

Thank you all for your replies. I appreciate your help. It would be helpful if the DJI Test engineers would weigh in to as how they tested RTH and came up with these numbers. I am not sure if we all agree that the RTH programing does not take the wind into consideration. For initial testing, I am going to say no. Taking that into consideration, the battery requirements to maintain 10 m/s would be in still air. This is for the computational display for the RTHI on the battery scale.

When you actually activate the RTH program the ball game may be different. It may try to maintain 10 m/s by decreasing pitch and increasing rotational power which would put a lot of inaccuracies into the RTHI. This is why I am asking these questions. This is a good discussion since this is the go to programing for problems.

You should be fine. DJI has a build in margin of error of 25% of the total battery capacity so it’s not going to take an additional 25% battery to RTH as long as the wind speed doesn’t exceed the 10 m/s limit. 10 m/s wind speed is a lot. It is uncomfortable to fly in those conditions anyway but even if you do you should give yourself extra time anyway.

I really wouldn’t rely on any of this. Best is get the experience to know when you need to come home based on voltage regardless of the calculated RTH time or even the battery percentage.

 

[DanMan32]

You are right but RTH is different. It uses a ground speed velocity instead of pitch angle. Try it, no matter the wind speed it will always RTH at 10 m/s ground speed unless wind exceeds 10 m/s. That’s why it says not to fly in winds exceeding 10 m/s and why you’ll get wind warnings to land immediately if you do

Then it isn't flying at 10m/s ground speed in that scenario, it is flying at 0. It will increase attack angle as much as possible to keep 10m/s ground speed until it reaches its max attack angle.

 

[jjckflsh]

You all bring up some great points. There needs to be a little more clarification as to what some definitions and indications mean. The whole reason I asked this question was to calculate some rules of thumb regarding winds when using the RTH feature. I my opinion, I think the 10 m/s gave the highest ground speed per unit of energy. Similar to best range airspeed. Not sure about this because max flight time is done at 25 kph which is much faster. Since there is no airspeed detector the drone is programmed to maintain a zero wind velocity of 10 m. Maybe because of the visual sensing system? I wonder if it uses current power consumption levels to compute energy requirements for the RTHI?
Also, what specifically does the high wind warnings mean and how are they computed? I am guessing that there is a power consumption level that triggers this? In addition, what does the Max wind speed resistance of 29 to 38 kph mean?
Jack

 

[DanMan32]

Wind warnings are probably governed by the attack angle needed to hover or move as commanded. If it exceeds a certain angle, warning appears. It tends to be quite conservative. Perhaps it should provide % of risk.

It may also consider rate of change. A heavy but rather steady wind probably has less risk than one that rapidly changes speed and/or direction.

 

[brett8883]

Then it isn't flying at 10m/s ground speed in that scenario, it is flying at 0. It will increase attack angle as much as possible to keep 10m/s ground speed until it reaches its max attack angle.

Close.

The drone can fly up to 20 m/s air speed so you’d need a wind of 20 m/s to make it fly at 0 into a head wind. In reality it can actually fly faster than this as well. Sport mode is limited to 35 degrees but the aircraft can pitch up to 45 degrees max which it will do if it must to maintain the 10 m/s RTH ground speed.

 

[brett8883]

Wind warnings are probably governed by the attack angle needed to hover or move as commanded. If it exceeds a certain angle, warning appears. It tends to be quite conservative. Perhaps it should provide % of risk.

It may also consider rate of change. A heavy but rather steady wind probably has less risk than one that rapidly changes speed and/or direction.

Yea this gets calculated and if it factors into the RTH calculation it’s unknown. So far there isn’t any evident that it does but that doesn’t mean it doesn’t it just means we need more testing.