This, and the previous reported incident, suggest that how ever much the developers of the Watchkeeper software tried to anticipate and build in safety features, auto-abort sequences etc, the damn humans always managed to interfere and stuff it up.
Army Watchkeeper drone flopped into tree because crew were gazing backwards
The latest British Army Watchkeeper drone crash happened because its crew became fixated on live footage from an onboard camera instead of their instruments, a Ministry of Defence report has revealed. An internal Defence Accident Investigation Branch (DAIB) report detailed why the semi-autonomous aircraft, tail number WK050 …
COMMENTS
-
-
-
-
-
-
Saturday 31st August 2019 08:54 GMT aberglas
Who was flying the airplane?
The human or the computer?
That was the problem. It should be one or the other. If the humans were not capable of doing a full landing they should not touch the controls. At all.
Semi automation is an issue for airliners too. They have things like Auto throttle that control the speed. Meaning bad pilots do not monitor air speed. So when it goes wrong, planes crash. (E.g. Ariana SFO a few years ago)
-
-
-
This post has been deleted by its author
-
Friday 30th August 2019 17:59 GMT Mark192
From the article: the camera... had been retracted to its stowed position, facing backwards and downwards
The camera showed that it had left the centreline of the runway and was now over grass. Their insructions for this situation, when landing, and was to cut the engine.
They weren't aware that it had aborted the landing and was attempting a go around.
-
-
-
-
-
Saturday 31st August 2019 15:50 GMT John Brown (no body)
"And no commenter has mentioned putting Royal Artillery troops in charge of a drone - the thing probably crashed because they saw a nice howitzer somewhere."
No one seems to have noticed that the RA guys were trainees, or at least under instruction since the article states they were under the supervision of Thales contractors. What were the "supervisors" doing?
-
-
-
-
Thursday 29th August 2019 11:59 GMT Peter2
Re: Aircraft 1: Humans 0
So the aircraft's systems decide to execute a go-round, while the wetware at the end of a joystick decides it knows better and causes expensive noises?
Well, this is the problem with having people with no flight training flying the things. To be fair, they did follow the right documented procedure. However, this is a perfect example of why people need training; in the time it takes to look it up (or google it) the "right" procedure in their manual was then used which would have probably been fine when the drone rolled onto the grass. (presumably because the reason "cut throttle if on grass" is a proceedure is that they know that the laser altimeter gives as bad readings on grass as an optical mouse with a hair under the red bit does?)
The problem was that by time they blindly implemented the documented procedure without understanding why it had been mandated the situation had moved on; the drone had started taking off again for a go around, and following the "cut throttle if on grass" procedure resulted in the crash.
Remediation to prevent occurrence appears pretty simple.
1) Give everybody really elementary training in the basics of flight.
2) Make them take off and land something in microsoft flight sim until they can do it without crashing, including go arounds when their approach is a bit dodgy.
3) Provide everybody flying one with a few hours real flight training with the Air Cadets Air Experience Flights. The cost of using an existing organisation with existing aircraft to train everybody flying things during the middle of the week when they have little use from the University Air Squadrons or for RAF basic flight training is going to be negligible, and almost certainly less than the purchase price of this crashed drone was.
4) Stick a cheap mobile phone camera in the front of the next version of drones to provide a wide angle "pilots view" of what's ahead.
-
-
Thursday 29th August 2019 12:32 GMT Anonymous Coward
Re: Aircraft 1: Humans 0
"To be fair, they did follow the right documented procedure."
The executed a procedure but my reading of it is they executed it too late - the correct procedure was to:
"The DAIB said: "The most appropriate [flight reference card] drill ...stated: 'If UA [unmanned aircraft] not maintaining centreline axis: Engine cut……..Command'.""
I'm assuming that once they saw grass, it was clearly not maintaining the centre line and they hadn't been looking at the correct part of the screen to view deviation from the centre line or had their hand in a position to react if they were following this procedure. It is unclear how much further delay there was between the power being cut - while it drifted into trees, I am assuming that it had gained some height and speed since executing the go around which may have resulted in a significant distance being covered while unpowered (as in hundreds of meters).
In terms of training, operators get to set way points or cut power - they don't have the ability to alter the direction or power level so a forward facing camera or training with flight simulators is unlikely to be helpful.
Removing the ability to see images from the stowed camera would appear to be helpful to ensure operators avoided distraction.
Improved operator warnings for when the craft is executing an unexpected or emergency procedure to allow operators to realize and adjust their response may also be useful.
-
Thursday 29th August 2019 16:09 GMT Peter2
Re: Aircraft 1: Humans 0
In terms of training, operators get to set way points or cut power - they don't have the ability to alter the direction or power level so a forward facing camera or training with flight simulators is unlikely to be helpful.
Well that's another design flaw there then since the rules of the air require powered aircraft to get out of the way of unpowered aircraft such as gliders and hot air balloons that have little to no ability to evade, as well as getting out the way of airliners etc. There definitely should be provision for human supervisors to assert flight control when and where appropriate.
-
-
-
-
-
Friday 30th August 2019 09:40 GMT Peter2
Re: @Peter2 -- Aircraft 1: Humans 0
have yet to deploy to somewhere other than their training base.
-
-
-
-
This post has been deleted by its author
-
Friday 30th August 2019 08:27 GMT NeilPost
Re: Aircraft 1: Humans 0
Being fair fully qualified and trained pilots could not fly a MCAS’d 737max that some ‘too clever by half’ engineer had cooked up.... to address airworthiness/fly ability issues that come with a passenger place of petite dimensions with comedy-tits sized engines... instead of designing from scratch around the far larger size of next Gen engines.
-
Friday 30th August 2019 09:33 GMT Intractable Potsherd
Re: Aircraft 1: Humans 0
@ Peter 2: "... presumably because the reason "cut throttle if on grass" is a proceedure is that they know that the laser altimeter gives as bad readings on grass as an optical mouse with a hair under the red bit does?
Why have such a complicated system when "weight on wheels" would be definitive?
-
Friday 30th August 2019 13:30 GMT Mr Sceptical
Re: Aircraft 1: Humans 0 - real pilots aren't any better!
I think you'll find a trained pilot is likely to be worse than a button pusher!
https://www.theregister.co.uk/2016/08/12/watchkeeper_drone_wk031_crash_report_1m_uav_destroyed/
I'm sure I read on The Reg another article about the loss rates of US drones flown by actual pilots being higher when they land manually (theirs are 'flyable' in the manned aircraft sense) than under computer control.
Basically, Lord Flashheart types think they can land better than the computer, only to be proved wrong when they write them off... wooof!
-
-
-
-
-
-
Friday 30th August 2019 09:43 GMT Intractable Potsherd
I was in a plane that landed very similarly to that on my only flight to Brussels - a Bombardier turbo-prop plane. The wind was terrible, and I could see the runway stretching out in front of me as I looked through the window over the wing! The gust that made it land on only one leg was the only really scary bit!
-
Friday 30th August 2019 17:18 GMT Orv
Crosswind procedures vary from aircraft to aircraft in interesting ways.
The basic problem is the wind is moving sideways, you're moving relative to the wind, but you have to line up with a runway on the ground.
High-wing aircraft, and most small planes, will usually side-slip into the wind -- they keep the upwind wing low, which causes the airplane to fly somewhat sideways, while still keeping the nose pointed down the runway. Done properly this results in initial touchdown on the upwind main gear wheel.
Large jetliners can't do this because their low-mounted wings and engine nacelles would drag on the ground at any significant bank angle. So they crab into the wind, and usually touch down still in a crab. The main gear is behind the center of gravity, so once it makes contact the plane will tend to naturally pivot into the wind; in the video you can see the pilot helps it along with some right rudder. Part of certification testing for a new aircraft is proving the landing gear can handle these kinds of side loads.
Sailplanes (which I used to fly) have a similar problem -- they have extremely long wingspans for their size, 15 meters or more, so side-slipping would result in a wingtip touching down first and the aircraft cartwheeling. We would make our approach in a crab, and then just before touchdown we'd kick in a bunch of rudder to de-crab. I found this an immensely satisfying maneuver -- it happens within a few feet of the ground, so there's an intense sense of motion through 3 dimensions, and you can see the aircraft respond to your actions in an immedate way that you normally don't. Most pilots hated practicing crosswind landings but I thought they were a hoot.
-
-
-
-
Thursday 29th August 2019 12:41 GMT Graham Cunningham
Beautifully done.
Because I love it so much, here's the Steve Lyons compilation from many years ago set to Enigma's Return to Innocence:
https://www.youtube.com/watch?v=9VCC3SaJECE
And one of any number of Medeira videos for anyone who likes their holiday to start with a white-knuckle ride:
https://www.youtube.com/watch?v=zkAFUazoFbM
-
Thursday 29th August 2019 14:59 GMT MiguelC
Re: Beautifully done.
Regarding the second video, let me tell you that Madeira's airport is fairly safe now comparing to what it used to be before its runway expansion from 1800m to the current 2780m. It was quite (in)famous among pilots for its extremely rough conditions and passengers used to brake with their feet as the plane touched down (the breaking was really hard then). Another brilliant thing was (and still is, to a point) that as part of the runway is built over the sea shore all you see during the approach phase, as a passenger, is water all around… you really need to trust the pilot!
-
Friday 30th August 2019 08:18 GMT Pen-y-gors
Re: Beautifully done.
Thinking of fun airports...
Who misses the old Kai Tak airport in Hong Kong? Another one extending out into the sea, with a flight path descending between blocks of flats on each side - you could wave to people putting out their washing next to you!
And Alderney - tiny airstrip that starts on the edge of a cliff several hundred feet high! You fly in and are convinced you'll hit the cliff.
-
Friday 30th August 2019 10:12 GMT Caver_Dave
Re: Beautifully done.
Gibraltar is another interesting place to land - both ends of the runway are in the sea.
When I last went, we weren't allowed into Spanish airspace, so it was effectively a dive towards the East of Gibraltar and then turn sharp right at the very last moment. Very squeaky bum, and very hard braking!
-
-
-
-
-
Friday 30th August 2019 09:53 GMT Anonymous Coward
Observers, however, are not supposed to crash into things. There's two basic jobs in the RA: Firing big lumps of exploding metal at things, and identifying the position and nature of the things. For the latter, all kinds of things have been used, from hole in the ground to balloons and aircraft.
In a real war the life of a drone would, I suspect, be extremely short. Landing could be just something that the few that survive the first mission need to do.
-
-
Monday 9th September 2019 11:09 GMT Anonymous Coward
£15 million actually over all costs. It is hard to imagine that a half tonne vehicle with a Wankel engine could cost that much but hey, it's the Israelis and Thales, those illegal settlements don't build themselves. I suspect if we'd gone to a Chinese vendor we could have got them for under 6 figures each, and the success rate wouldn't be any worse. Synthetic aperture radars got an awful lot cheaper when people started making them using commercial rather than military electronic fabrication technology.
-
-
-
-
-
Thursday 29th August 2019 11:43 GMT Chris G
It sounds to me that two things (at least) need to change.
First , the aircraft is autonomous except for input at navigation waypoints, meaning the operator's attention will never be as good as that of someone who is actually flying it remotely so this kind of distracted operator situation is a result of the system.
Second the 'wheels on ground' indicator is over complicated and would function better and more reliably my opinion with an analogue switching system. Somewhat cheaper and easier to maintain too but I guess that is not in Thales interest.
-
Friday 30th August 2019 07:50 GMT YetAnotherLocksmith
It's the worst of both worlds, and will be the cause of many a death with the soon to arrive "nearly self driving cars".
Imagine travelling for 7 hours from Edinburgh to Birmingham, maintaining awareness of everyone on the road, but without actually changing speed (you're locked at 70, or, mostly, 50 or even 30!), changing lanes, keeping a safe distance - all that's done by the car now. But you've got to be ready, in a heartbeat, to "take over" and save the day off the computer cocks it up, and, say, swerves into contraflow traffic at 50mph. No chance!
Same with this aircraft. It's boring, and you're a "passenger".
-
Friday 30th August 2019 17:22 GMT Orv
I do wonder why they didn't use a WoW switch. I suppose on rough landing sites it might be unreliable, since a lightweight drone would bounce over bumps a lot. It's also not clear to me how much suspension the drone actually has; a WoW kinda needs some kind of compression element, like an oleo strut or similar.
-
-
Thursday 29th August 2019 12:09 GMT Ima Ballsy
And then a few days latter the email arrives:
" am well aware lwatchkeeper ŏne ŏf your pass. Lets get right to the purpose. No ŏne has paīd me to invėstīgatė àbout you. You do not know me and you're probàbly thinking why yoũ're gėtting thīs email?
ī plàced a software on the X vids (pornŏgraphy) web-sīte and guess what, yŏu visitėd thīs website tŏ hàve fun (yŏũ know whàt ī mėan). When yŏu were watching video clips, your internėt brŏwsėr started ŏũt working as a Remŏte control Desktop with a key logger which prŏvided mė access to yoũr display screen and càm. Right after that, my software obtained your complete contacts from yoũr Messenger, sŏcīal networks, and e-maīlaccount. àfter that i màde à doublė vīdeo. Fīrst part dīsplays the video yŏu were viewing (yŏũ have a good taste lmao), and second pàrt shows the view of your wėb camera, yea its you.
You wīll have a pàir ŏf choīces. Wė wīll rėàd each of these choīces in details:
First alternative is to disregard thīs e-mail. Consėquently, i wīll send your video tŏ just àbout àll of yoũr cŏntàcts ànd thus think abŏũt regarding thė awkwardness you will sėe. Not tŏ forget if yŏũ are in a loving relàtiŏnshīp, how it wŏuld affect?
Lattėr solutīŏn shoũld be to compensate me $1346 USD. i will call it à donation. in thīs case, i wīll withoũt delay eliminatė yoũr videotape. You can go on yŏur wày of life like thīs never tŏŏk plàce and you never wīll hear bàck again from me.
Yŏũ wīll màkė thė pàymėnt via Bitcŏin (if yŏu don't know this, search for 'how to bũy bitcoin' īn Google).
BTC àddress:
13jz1hv4pXd6dd5JJd9rXqtLbgfZxQGCte
[CaSe SėNSiTiVe copy and paste it]
in case yoũ are thinking of gŏing to the law, look, this email message can not be traced back to mė. i hàve deàlt wīth my actions. ī am just not àttemptīng tŏ chargė a feė a huge àmount, i just want to be pàīd for. Yŏu now have 48 hours in order to màke the payment. i'vė a specific pixėl īn this e maīl, ànd at this moment i knŏw that yŏu have read thīs emaīl. īf i dŏ nŏt get the BītCoins, i wīll definàtėly sėnd ŏũt your video to all of your cŏntacts īnclũding members of yoũr famīly, co-workėrs, ànd màny ŏthers. Howevėr, if i do get paīd, ī'll destroy the recording immedīatėly. This is the nonnegotiàblė offer, and thus do not waste my time and yours by replying to this email. if you really wànt proof, rėply with Yėà! then i wīll sėnd your video rėcording tŏ yŏur 11 friėnds. "
-
-
Friday 30th August 2019 07:52 GMT David 132
Re: Starglider was better!
Yeah! Where’s the Fluent/Modern design elements? Why isn’t the text cleanly rendered in a super-thin font in pale grey on a blinding white background? Why aren’t all the nerdy wonkish options that no-one cares about (“trim”,”waypoint”,”speed” etc...boooring) hidden behind a hamburger menu or a simple, intuitive swipe-left-then-across-at-precisely-78° gesture? Where, I ask you, is the social media integration?? I bet it doesn’t even support colour emoji, let alone animated ones.
Bloody defence contractors have NO idea how to do software.
-
-
Saturday 31st August 2019 16:36 GMT John Brown (no body)
Re: Starglider was better!
And not forgetting that it only works with a remote network connection so "important" live updates can be delivered as soon as possible, either completely changing the entire UI or, worse, leaving the UI almost but not quite exactly the same, just the up and down buttons have been swapped due to user request (but not the request of the users who actually use it)
-
-
-
Thursday 29th August 2019 12:35 GMT Chrissy
From a different article, same subject, similar failure:
Said this before, I'll say it again:
Surely it would be a net saving by sending the "commander" of each single drone's crew to get their PPL - £10k - or even a gliding "get to solo" course - £2k-ish - so they have an understanding of the basics of flight and why an aircraft may be landing long or just doing something that, to the untrained eye looks a bit weird but is just the software doing its thing, as - you'd hope - somewhere along the line a pilot has been involved in its development.
Yes, it has no actual "controls", but it will at least stop most of these "I can't do a lot, but I'll do what the system allows me...... Oops" situations.
With 50 aircraft costing £800m, this single crash "cost" £16m; at that price you may as well send anyone in the entire Army who will ever have anything operationally to do with a drone on a gliding course and still have plenty of money left over.
-
This post has been deleted by its author
-
-
Friday 30th August 2019 08:25 GMT Pen-y-gors
Re: This was a good landing
I have a lovely grainy old photo from about 1917, Bi-plane nose down in the ground, three men in uniform standing next to it. One with an arrow and 'X'. X was my grandpa, who was an RFC flying instructor at Brooklands (and it seems likely he taught the future George VI to fly). I believe it was a cabbage field. I assume it was the pupil who was in control.
That was a good landing.
-
Friday 30th August 2019 17:27 GMT Orv
Re: This was a good landing
Those WWI-era planes were absurdly difficult to land. Narrow landing gear, no brakes, the suspension was often just bungee cords. I've gotten the impression minor (and sometimes not so minor) landing mishaps were pretty common.
I remember a discussion on an aviation forum about what types of farm fields were worst to land on. One guy declared with some authority that the worst was soybeans. Apparently he was a cropdusting pilot and had had an AgCat nose over while trying to make an emergency landing in a soybean field. He compared it to landing on velcro.
-
-
-
Thursday 29th August 2019 13:20 GMT Anonymous Coward
Right response
> the operators' quick-reference cards stated that the right response to a Watchkeeper rolling off the runway centreline was to cut the throttle.
It's not clear whether this applies to a Watchkeeper that has already landed rather than one that is just about to touch down. In the latter case hitting the 'go around' would seem to be a better initial choice. There must be instructions to the operators on when to trigger a go around so it would be useful to see those.
Secondly, as we are told that the camera was part of the payload, and not part of the operator's control screens - how are they supposed to see whether the thing is lined-up or not, rolling off the centreline or not? Do they have someone in the control tower with binoculars who radios through what's happening? Where was this person during this incident? More questions, as always.
-
Friday 30th August 2019 07:58 GMT YetAnotherLocksmith
Re: Right response
Not really.
The drone came close to landing, but decided it was too far off the centre of the runaway to land, due to the crosswind. It therefore powered up again to go around and try again.
Alas, the "crew" saw the runway turn to grass on the camera display, panicked, and cut the power. They thought it was on the ground, because they were looking at the grass on the video stream, instead of the altitude. Without power, the drone flew a short distance more and hit a tree.
-
-
Thursday 29th August 2019 14:29 GMT Danny 2
Thales Kill Whales
Oui, I know now the pronunciation doesn't rhyme. Thall Ess. A deaf whale is a dead whale and Low Frequency Active SONAR kills whales wasn't a great slogan. My idea was to fill two six foot inflatable plastic whales with helium, cover them with rape alarms, and leave them tethered outside a Thales office. It/they didn't float, too heavy for the helium.
Now I've got a six foot inflatable whale in my four foot bath, and a wee lass in Dumfries has, god knows, she is a grown woman now.
I met a single mother peace protestor on the beach at Coulport playing with her two toddlers when I was trying to unload this six foot inflatable whale so I said, "I've got a six foot inflatable whale if you want it."
Her response was quite funny. "Oh, I bet you say that to all the girls."
-
-
Friday 30th August 2019 08:57 GMT bpfh
But I guess the drone drivers did not get IFR training...
One way is driving the vehicle with a hood over your eyes allowing you only to read the instruments and not look out of the windows. I’m not a pilot (but still a plane geek), and had the chance in the mid 90’s to “play” on a French Army basic instrument training helicopter simulator. Not much more than a cut down version of the DOS game “LHX” tied into a real Allouette 3 cockpit and controls. In the original game the ground surfaces had the same colour and if I remember were flat, but the closer you got to the ground, dots would appear regularly spaced as a grid over the ground, giving some impression of texture and perspective: the closer to the ground you got, the wider the spacing between the dots became. The army helicopter sim did not have this, and threw in simulated terrain hazards. I made 3 consecutive Controlled Flight Into Terrain as I was watching the screen and not the instruments where the radar altimeter (yes it had one, probably just for this) was slowly winding down to zero as I was in straight and (low) level flight and flew the sim into the side of a hill. Lesson then learned from the flight instructor running the sim: don’t trust your eyes, Always trust your instruments.
Icon => what my alouette looked like after flying by the seat of my pants...
-
-
Thursday 29th August 2019 17:02 GMT Milton
Situational awareness
This wouldn't be the first plane crash caused, or made worse, by slavish adherence to the check list, so I hope these guys' Interview-Without-Coffee wasn't too much of a career-killer: it sounds an awful lot like they were committing a mistake that many good professional pilots have made down the years, instinctively trying to maintain situational awareness visually rather than by watching the instruments. It is very, very, very hard indeed to train would-be pilots to ignore their eyes and ears and to rely on the instruments. At least this time no one died.
But it does raise an interesting question, to which I don't know the answer: how do you maintain situational awareness when flying a big drone? You really do have to rely upon instruments, because there are no underpants wedged into a cockpit, but there are cameras to seduce you into believing you are right there ... I would imagine that a flight psychologist could write lengthy dissertations on the differences between training actual in-plane pilots and the remote drone kind. Some of the distinctions must be amazingly subtle yet important.
-
Thursday 29th August 2019 20:39 GMT Anonymous Coward
Re: Situational awareness
I recall being told that in the 1950s experiments were done on injecting pilots backsides with Novocaine and then getting them to fly a plane. They got into difficulties.
The conclusion was that the information fed to you by your backside is exceedingly important. And of course totally missing in static flight simulators.
-
Friday 30th August 2019 12:06 GMT Spamfast
Re: Situational awareness
it sounds an awful lot like they were committing a mistake that many good professional pilots have made down the years, instinctively trying to maintain situational awareness visually
No, read the article. They were both watching the video surveillance package monitor showing a view backwards out of the plane instead of the operator control monitor. They were not supposed to be doing that. They then panicked and did something without checking the instruments which resulted in the loss of sixteen million pounds of aircraft plus however much it cost for the investigation - of our money.
So they were goofing off to start with, then they lost discipline and finally failed to follow procedure. I'd court marshal them both for dereliction of duty and hopefully they'd get thrown out with a dishonourable discharge after a goodly spell in the glasshouse.
Also, what the hell were the guys from Thales that were supervising them doing? Probably also watching the pretty art video on the other monitor as well - again, which was not why they were there. That's grounds for demanded a replacement aircraft free of charge in my book. (But then the MoD never actually makes contracts with the pond life arms companies that hold them accountable, do they? Otherwise the brass & civil servants who sign it off wouldn't get their bribes - sorry - non-exec directorships, fact-finding trips to the Bahamas, etc.)
On a wider point, why are we paying for all this tech if it's never actually used? Wouldn't the money be better spent on, oh I dunno, proper armoured vehicles for the lads & lasses to protect them from IEDs instead of giving them Land Rovers with a bit of extra tin riveted on?
-
-
Friday 30th August 2019 05:42 GMT Bob.
From my reading of the el Reg summary, the drone WAS [at least partially] on the grass.
Do pilots on manned flights perform go-arounds from the grass? Hazards are there! Ruts, rabbit holes, signage.Wheels dig into soft ground, throw clods of mud into the engines, etc.
I suggest the drone operators did exactly the right thing, albeit from non-standard input, and are being scapegoated.
-
Friday 30th August 2019 08:06 GMT YetAnotherLocksmith
No? It never touched the ground. If it had, the thing would've landed. The inertia of spinning up the wheels would've slowed it dramatically, as designed, and the aircraft, suddenly on grass, likely couldn't have powered away without serious effort, instead completing the landing.
It got *really close* to landing. Close enough to fool the "pilots".
-
Friday 30th August 2019 17:29 GMT Orv
Pilots on manned flights will initiate a go-around if they see they're going to land off the runway. The drone actually attempted to do the same.
That is, unless they're glider pilots. In that case a go-around is not an option. But in that case they were probably trying to land on grass to start with.
-
-
Friday 30th August 2019 06:33 GMT Richard 12
That UI is bloody terrible
What other events cause that bar to go red?
And why is the red so dark that you can't read the message?
A big red bar is going to instinctively cause people to hit the big red button. If the operators usually do nothing during landing, and the right action to take for an automatic TOGA is "nothing", then don't flash up a big red message for "DANGER!"
Real pilot warnings are deliberately made as different as possible. DON'T SINK, STALL etc
I can't work out what that square is trying to show (it's marked in "+/-1.5 deg" and not labelled), but there doesn't appear to be an artificial horizon or navball, velocity vector indicator, throttle indicator, or even vertical speed!
HTF are you supposed to have any idea what the aircraft is doing? They only appear to be told position over ground, everything else is numbers.
Sure, the operators aren't "flying" as such, but it looks like there's no way they could have much situational awareness, even under perfect conditions. How could the operator decide whether to abort, and which abort?
-
Friday 30th August 2019 08:11 GMT YetAnotherLocksmith
A heck of a lot of these commentards miss this vital paragraph:
"According to the DAIB, WK050 "failed to register ground contact during the ground touch window and auto-aborted", meaning it throttled up and took off."
The drone never landed, and the "pilot" cut the power thinking it had, and was crashing.