Fly away yikes!

[Deleted member 103366]

I have added what I think is the distance from the homepoint, purple plot, to the above chart.

One of the reasons I am pursuing this is, not so long ago I experienced an RTH height of 875ft and this one is 807ft.
I did not intentionally set the RTH height to that, an RTH height of 40m would be high for me, and I had put the 875ft down to me accidentallly touching and dragging the RTH height slider whilst intentionally moving/increasing the distance limit.
But I find it a bit odd that our two RTH heights are fairly similar and seem unintemtional. However I do suppose it is possible it is pure coincidence.

When I queried the events and how to stop the ridiculous climb, part of Meta4's response was
"Here's another useful pointer from p14 in the manual

In Smart RTH and Low Battery RTH, the aircraft automatically ascends to the RTH altitude. If the aircraft is at an altitude of 65 ft (20 m) or higher and has not yet reached the RTH altitude, the throttle stick can be moved to stop the aircraft from ascending. The aircraft will fly directly to the Home Point at its current altitude "

In my case I saw that the RTH height was 875ft and the max ceiling was well above that because earlier in the flight I had had the drone climb a 400ft cliff and fly over the top of it so I had given myself plenty of ground clearance. That said the RTH ceiling is capped by the max height ceiling so two things sort of had to be wrong to send my mini up there.

 

[Phantomrain.org]

Magnetic interference can send a drone anywhere the wind is blowing once it goes to Atti Mode and if there our Yaw errors any control is gone thus the term fly away.
This was not a fly away, its was pilot error based on obvious Panic.

 

[Deleted member 103366]

Phantomrain, I see very short flashes of "GPS atti" but generally as it is changing from GoHome mode to something else.
If I missed something can you please tell me what parameter, by precise name, I need to look at?
I would also like to know what "height" or name I need to look at to see the max height ceiling?
The only unidentified constant height that I can find is "maxHeight" in "details" and at 245m it is below what I believe to be the set RTH height. That is contrary to messages that I have seen in the app regarding rth height so I am wondering if it is the max height reached during the flight

 

[Meta4]

Well I went out to do a film of our local water tower and as I sent my Mac air 2 up to 200 ft it took off straight up until it disappeared. I tried all the controls to reduce height but no response from the drone. I hit the home button on the remote but still nothing. As I was about to give up I heard my drone close by and turned and saw it behind me above a tree about 100 ft high. I took control of it and landed it. Scared the life out of me, I saw $1000 flash before my eyes. I’m going to do a bunch of test flights before I try the water tower again.

Your flight data looks like this: DJI Flight Log Viewer - PhantomHelp.com

You had the drone up at 178 feet and initiated RTH.
As your RTH height was set to 245 metres (that's 804 feet), the drone started to ascend towards 804 ft.
You messed with the joysticks a bit and then cancelled RTh at 2:58.3 with the drone at 320 ft.
Then initiated RTH and cancelled, initiated RTH, messed with the joysticks a little as the drone climbed to RTH height of 804 ft when you finally cancelled RTH and brought the drone down manually.

It seems that you weren't aware of the height set for RTH or how to cancel RTH at any time to stop the acsent and bring it back down.

Magnetic interference can send a drone anywhere the wind is blowing once it goes to Atti Mode and if there our Yaw errors any control is gone thus the term fly away.

The drone never went into atti mode, never drifted anywhere and there's no hint of magnetic interference or any flying away.

 

[Fred Garvin]

You also busted the Toronto E700.....

 

[Phantomrain.org]

Your flight data looks like this: DJI Flight Log Viewer - PhantomHelp.com

You had the drone up at 178 feet and initiated RTH.
As your RTH height was set to 245 metres (that's 804 feet), the drone started to ascend towards 804 ft.
You messed with the joysticks a bit and then cancelled RTh at 2:58.3 with the drone at 320 ft.
Then initiated RTH and cancelled, initiated RTH, messed with the joysticks a little as the drone climbed to RTH height of 804 ft when you finally cancelled RTH and brought the drone down manually.

It seems that you weren't aware of the height set for RTH or how to cancel RTH at any time to stop the acsent and bring it back down.

The drone never went into atti mode, never drifted anywhere and there's no hint of magnetic interference or any flying away.

It was just a general statement I made about my experiences with magnetic interference .

 

[boblui]

The craft entered go home mode at 149 second into the flight. Immediately before that there was a "downlink restored" message indicating that the downlink has been broken. My best guess is that the trees in front of you was blocking the radio signals so the craft initiated RTH.

RTH was initiated at this position at a height of 54 meters.



Not sure if this picture is up-to-date but there were trees in front of you when looking towards the drone



After RTH was initiated, the craft ascended by itself as usual to the go-home height which was set to 246 meters. That's why you saw it going up.

In the RTH mode, throttle down control will only stop the ascent but it will not make the craft descend. That matches your report that the drone did not respond to control inputs.

Then you pressed the RTH button for multiple times resulting in RTH being cancelled and re-reinitiated repetitively. Fortunately the last press of the button put the craft into RTH mode and craft returned to the home point automatically.

If my guess about the tree blocking the radio signal is correct, you may want to pay more attention to ensuring LOS to the drone in the future. Good luck.

 

[InvisibleName]

You know, it would really help others if, when it’s demonstrated by analysis that a “fly away” didn’t actually happened, that the OP edited their thread title accordingly. E.g. “Fly way? Nope, I messed up”

 

[AMRE2ME2]

And once again, the drone did exactly what it was told to do.

 

[eipbroc]

Still odd that the connection was blocked by a few trees.

 

[Deleted member 103366]

As your RTH height was set to 245 metres (that's 804 feet), the drone started to ascend towards 804 ft.

Is 245 metres a typo Meta4?

I ask because I have just checked the logs of my flight mentioned previously and in that I know the RTH height and have a screen capture video of it. In the csv of the txt log of that flight the relevant "data field"? comes under the header "HOME.goHomeHeight [m]", (column GW or 205, if the csv of the log is opened as a spreadsheet).
HOME.goHomeHeight [m] is one of the entries in the chart I posted here and displays as 246m.

I know it's a small difference but as 245m is also a height I mention in post #23 in connection with another "data field" I think it could lead to confusion. If 245 metres is not a typo can I ask which "data field" you are looking at to get that value?


As an aside, in looking at the screen capture of the flight I mention I see that the "Max altutde" (as in the max height ceiling that can be changed in the app) was set to 283m and I can not find that height as a 'constant' in the spreadsheeted version of the csv of the txt log. It does appear in a few altitude entries in the spreadsheet but they seem to be in data fields of variable data.

So, I have a question, is the "Max altutde" included in the txt flight logs of the Fly app?

As a second aside and again stemming from my looking at the log of my flight " "maxHeight" in "details" " seems to be the maximum height reached during the flight

 

[Deleted member 103366]

You know, it would really help others if, when it’s demonstrated by analysis that a “fly away” didn’t actually happened, that the OP edited their thread title accordingly. E.g. “Fly way? Nope, I messed up”

Give them time, they haven't been logged in since 8.20 pm yesterday (according to the time stamps that I see in their profile, and I am not sure of which time zone is referenced)

 

[Repaid1]

Magnetic interference can send a drone anywhere the wind is blowing once it goes to Atti Mode and if there our Yaw errors any control is gone thus the term fly away.

The problem there is the term used. A aircraft that would do as you proposed would not be a fly away. A better term would be operator error and a "Float Away" combined with a "Flown Away". I am certainly not here to hash out things or point fingers. But the better half of new fliers and some even senior in their flying are using that term wrong and thus adding to confusion and DJI Scorn.

The actual term "Fly Away" is when the aircraft has a serious hardware failure and is not responding to either command from radio side or FC side and is basically dead and flips into a throttled motion off in some direction or retains last command. This was conceived way back with the Phantom 2 series and gave DJI the coined term of being habitual to this. This simply isn't the case.

However since the onset of the P3 series true "Fly Aways" are a thing of the past and well below a percent of a percent from my observations. Hardware failure will happen, but are EXTREMELY rare of this nature. Most would result in a plumetting aircraft, not a true Fly Away.

In your instance of situation. Magnetic interference would have nothing to do with a aircraft losing GPS lock which would be why it would go to Atti Mode. If the aircraft would encounter Mag errors in flight it would inhibit a true RTH. That being said it would not induce a auto RTH. IF in the event it could happen then the aircraft would rise above any interference to RTH normally anyway.

There are plenty of videos shown that most DJI aircraft are immune to radio towers, transformers, microwave transmissions or any number of myths and legends that people come up with these days. Simply the aircraft have redundant systems in many instances and "know" what to do had these systems fail, unfortunately the same can't be said of the latest (Year or so) operators (Nothing Implied).

Let's PLEASE, stop using the term flyaway unless it has been proven, DJI gets enough flack in the press and media and the haters groups...let's not add to the mix.

 

[boblui]

.... The actual term "Fly Away" is when the aircraft has a serious hardware failure and is not responding to either command from radio side or FC side ....

... A better term would be operator error and a "Float Away" combined with a "Flown Away" ...

Just wondering who defined this term ? besides "fly away" and "float away", how many other kinds of "away" are there in your dictionary ?

 

[itsneedtokno]

Just wondering who defined this term ? besides "fly away" and "float away", how many other kinds of "away" are there in your dictionary ?

Look at my thread in P&V...

There's "Blown Away" too. ?

 

[Repaid1]

Just wondering who defined this term ? besides "fly away" and "float away", how many other kinds of "away" are there in your dictionary ?

Don't really know other than been flying R/C for about 35 years and they were in place back then. The advent of quads with the commercially available P1 & P2 of course defined some by those that flew them. I have only been flying Quads since the advent of LiPo's, am a Nitro flier by nature before started with Heli's in Japan long before electronics (Gyros, IMU's Ect.).

Started flying Quads with a Walkera Scout X4 in the P2 era, and that was "Fly Away" central times. As I said I bought a P3P shortly after and have followed about every thread since on the biggest R/C sites with over 10,000 posts, so while new to posting here, I am known. I merely tried to correct verbiage (Fly Away), wasn't pointing fingers just trying to correct what I have known correct for years, YMMV... I imagine Google would be your best avenue had you want to know exact time lines and more terms along those lines.

 

[Meta4]

Is 245 metres a typo Meta4?

I ask because I have just checked the logs of my flight mentioned previously and in that I know the RTH height and have a screen capture video of it. In the csv of the txt log of that flight the relevant "data field"? comes under the header "HOME.goHomeHeight [m]", (column GW or 205, if the csv of the log is opened as a spreadsheet).
HOME.goHomeHeight [m] is one of the entries in the chart I posted here and displays as 246m.

Normally I'd check the set RTH height in the verbose CSV but there's something weird about the file and despite trying 4 times on two computers, all I get is the first 7 lines of the spreadsheet with no data on the set RTH height (or much else).
So I just used the height that the climb to RTH topped out at which was 245.1 metres

 

[Meta4]

To clear up some minor details ..

I sent my Mac air 2 up to 200 ft it took off straight up until it disappeared.
As I was about to give up I heard my drone close by and turned and saw it behind me above a tree about 100 ft high. I took control of it and landed it.

The craft entered go home mode at 149 second into the flight. Immediately before that there was a "downlink restored" message indicating that the downlink has been broken.
My best guess is that the trees in front of you was blocking the radio signals so the craft initiated RTH.

The downlink restored message was at 2:16.9 after a blank of 0.9 seconds
RTH initiated later on at 2:29 with no lost time before it which suggest that RTH was initiated manually by the OP.

Fortunately the last press of the button put the craft into RTH mode and craft returned to the home point automatically.

The OP cancelled RTH at 4:45.1 and 4:47.8 he pulled the left stick hard down and brought the drone all the way down manually.

 

[Meta4]

Just wondering who defined this term ? besides "fly away" and "float away", how many other kinds of "away" are there in your dictionary ?

Flyaway (with respect to drones) is a term that was in use before April 2015 and it meant "I lost my drone and don't understand why".
After the release of the Phantom 3 with the DJI Go app to record flight data, we were able to investigate flight incidents and discover what happened, why the drone was lost and often, where it might have come to rest.

The term should have been retired years ago because what we've learned from reading the data of hundreds of flight incidents, is that drones don't "fly away" at all..

 

[Starz]

that a “fly away” didn’t actually happened,

Or, when we see the words “fly away”, we could just skip on by knowing that it really didn’t happen.
But the reading material is entertaining thanks to Meta4.