General drone intelligence

[Yorkshire_Pud]

He was lucky that it completed safely, but anything could have happened.

Why lucky?
Was that not the intended behaviour when using that App?

 

[Meta4]

It's good to know it returned to home when you requested it. i was talking to someone yesterday that used a 3rd party app automated flight. The drone lost contact but still flew the mission out of range. Took the required pictures with no contact and flew back. He was lucky that it completed safely, but anything could have happened.

That's a plain vanilla operation for Litchi, not an anything can happen situation..
The drone saves a homepoint as soon as it gets good GPS location data, the waypoint mission can be programmed to continue if signal is lost, and the drone returns to the home point when the mission is finished.

 

[BetterBizWorks]

Why lucky?
Was that not the intended behaviour when using that App?

If the radio disconnects from the drone it should return to home for safety. Some could argue that it should complete the mission, but the pilot need to be able to control it to stop an accident from happening.

 

[Zbip57]

If the radio disconnects from the drone it should return to home for safety. Some could argue that it should complete the mission, but the pilot need to be able to control it to stop an accident from happening.

Similar to the options normally available for loss-of-signal response (RTH, Hover, or Land), third party waypoint apps like Litchi offer an option to Continue with Mission upon loss of control signal. The default setting is usually set to Return-to-Home in both cases.

If the drone continued a waypoint mission after a loss of control signal, it's because the pilot deliberately chose and selected that option. The pilot should not have been surprised by this configured behaviour.

That said, not all drones are capable of continuing waypoint missions upon loss of signal. My old Phantom 3 Pro is able to do that because the waypoint mission is uploaded and stored onboard. My Mavic Mini is unable because it hasn't got the onboard storage required for that. Instead Litchi waypoint missions are steered and directed by the app itself, which requires a constant control signal connection. Upon loss of control signal, the Mini will perform its configured loss-of-signal option (RTH, Hover, or Land).

 

[Yorkshire_Pud]

If the radio disconnects from the drone it should return to home for safety. Some could argue that it should complete the mission, but the pilot need to be able to control it to stop an accident from happening.

Not necessarily, I believe that with some apps they are designed to carry out the programmed flight and then RTH. It's never interested me so I don't pay that much attention to waypoints but I do recollect someone saying that at least one app is designed that way and presumably writes the instructions into the drone's memory.

 

[BigAl07]

If the radio disconnects from the drone it should return to home for safety. Some could argue that it should complete the mission, but the pilot need to be able to control it to stop an accident from happening.

That's not how Litchi (and several other Way Point Flight Apps) are written. It's not a flaw but a FEATURE built into the aircraft and it's useful in many instances.

Keep in mind that many of these 3rd Party Apps are written to survey/cover large expanses of distances and could be used in areas not covered by FAA regulations (or flown with Waivers).

Let's not paint with such a broad brush

 

[hedbonker]

Well like I said my question was more about the built in waypoint mission system for modern DJI drones. Not so much RTH. I use Map Pilot Pro and I have had the drone disconnect during the mission a number of times without incident. - The drone finishes the mapping mission and does a successful RTH. Now I would not trust Map Pilot Pro to prevent the drone from going to some location 20 miles away or something in the off chance I launched the wrong mission - It's HIGHLY unlikely for that to happen but I was just curious about how the DJI software might behave.

 

[Jamesz]

@hedbonker the issue you mentioned about moving 100 miles away ,was primarily a compass related thing ,and had nothing to do with the actual homepoint directly
with todays drones, if on start up the drone senses an anomily, between what the satellite information and the compass seem to be saying,then you will get a compass calibration required warning on the screen
what can happen sometimes ,is that the drone has been launched, before enough satellites have been locked , and at some point away from the take off point ,the satellite lock is established,then that point will be where the drone will return to
the two messages saying homepoint is set ,once on the ground ,and then as the drone takes off, are just to make sure, that the operator has got the message,the second one is a comfirmation of the first, the homepoint remaines the same
as far as a homepoint mission is concerned ,then realistically once the drone has located its compass heading and position by GPS coordinates ,then it should realise that the stored waypoint mission coordinates, and the actual coordinates ,are not the same ,and maybe abort the mission from taking place, whether this is the case or not i do not know ,just summising
maybe someone can try this out and see if it is the case

I think you are incorrect about the gps/compass at start up. Gps does not provide orientation guidance when static, instead the compass gives the IMU a directional reference. Once airborne, the IMU uses that to provide heading orientation with the compass secondary. The need for calibration comes from anomalous values generated by local interference, ie nearby metal. There was never a need to recalibrate in a new location and certainly not before each flight.
Checking the drone's actual orientation matches that indicated on the map at start up is the best way to avoid compass related accidents.

 

[old man mavic]

@Jamesz for sure when the drone is aquireing sats and just sitting on the ground ,if there is some metal object close by, or indeed under the ground such as rebar ,then it will give the compass an incorrect reading ,on older drones this often resulted in toilet bowling just after take off ,and also with older drones, then a compass calibration was often called for, if the drone had been moved more than 50 miles, from its previous take off point
as the sophistication of the drones internal systems has improved ,with advances in the algorithms used to run the flight controller, this need has been almost eliminated

 

[Droning on and on...]

I'm pretty sure the DJI firmware doesn't ignore low-battery RTH even if executing a waypoint mission with continue on disconnect (Failsafe RTH).

Anyone know for sure? A brief look at a few user manuals for different models that support autonomous operation didn't find anything.

 

[Droning on and on...]

There was never a need to recalibrate in a new location and certainly not before each flight.

Something not well-known by the general public – most of whom never use compasses for anything serious – is that magnetic north is not true north. This is called Magnetic Declination, and is very important for navigation.

GPS, which our drones use for navigation, is based on true north. So it's important that magnetic declination be corrected for.

Depending on longitude, MD varies from 0–30° in N. America:

Most of the time late model DJI drones handle compensating for MD automatically once they get a fix and know their longitude. The data in the map above is contained in the firmware. MD varies over years, so accuracy depends on updating this information from time to time. The farther north or south, the more often this data needs to be updated to stay accurate. I think some older Phantoms did not correct for MD automatically, so always called for a calibration when moved far enough.

Even still, a large change in MD can occasionally cause a late-model drone to call for a compass calibration, if it just can't make sense of the compass data within whatever error tolerance the engineers built in.

Also large relocation north or south can cause the compass to get confused because as you get close to the magnetic poles, the magnetic field vector starts to have a significant Normal component in addition to the Tangent component (i.e. it starts to point toward the surface), which must be ignored, and calibration may be necessary to deal with this.