There has been a lot of praise from the aviation and mainstream media (and some inaccurate reporting, too) for the tech behind the first Garmin Autoland activation that successfully planted a Beech King Air back on the runway in Colorado when its pilots couldn’t.
But we think there hasn’t been enough focus on the good work accomplished by the shop—Blackhawk Aerospace—that installed this first aftermarket Autoland and Autothrottle interface in this King Air. Moreover, because a real-world Autoland activation creates an in-flight emergency as this deployment did, shops can’t fully test the Autoland before handing the airplane back to the pilots, though there are plenty of tests that are performed. Want to know if it will really work? Press the button and try it for real.
In this episode of the Smart Aviator podcast, we sat down with Blackhawk’s Conrad Theisen for a detailed look at the installation and the sequence of events during a real-world Autoland deployment.
Good video interview of the Garmin Autoland airframe works, but it would help to be clear about its operational limits.
From what I read in the Garmin and OEM material, Emergency Autoland doesn’t just pick any runway. It wants a suitable RNAV (GPS) approach with vertical guidance, enough runway, acceptable terrain, usable weather, and reasonable fuel margins. That leaves out a lot of VFR-only airports and places without that kind of GPS approach, even though we use them all the time.
This incident also started in cruise with plenty of altitude, which is about as good as it gets. Many real problems don’t start there. A depressurization at altitude is one thing; a bird strike, pilot medical issue, or partial-panel problem is something else. In those cases the system has less time, fewer “Autoland-approved” airports in range, and may have to climb just to find one that meets its own rules.
None of this is a knock on Garmin; the system did what it was designed to do here. I’d just like to see the limits spelled out clearly,what qualifies as an Autoland runway, what does not, how the logic behaves when you’re not up high and stable, which airports in the database are actually Autoland-capable, and how weather is weighted in those choices. That’s the kind of detail working pilots and owners can use.
Damn, Raf … here I am at 0300 pounding away on the keyboard to chide Larry and when I pressed “post” it says, “2 Comments.” HUH ??? I didn’t see another comment until I pressed refresh. OH! Raf is out west pounding away on his keyboard, too. 😃
OK … good comment. I forgive you for what you wrote the first time.
Glad to hear from you, Larry.
“Want to know if it will really work? Press the button and try it for real.”
NOOOOOOOOOoooooo !!! Don’t say that to a pilot of an ‘equipped’ airplane!
In the other AvBrief article, I already predicted that would start happening … don’t pour gas onto a fire here, Larry. Unless there’s far more to the situation that evolved in CO which hasn’t yet been revealed, I am of the opinion that two supposedly qualified 135 pilots took this pressurization / autoland situation WAAAY too far … to include nonchalantly exiting the airplane when the props were still turning and it was sitting on the runway safe. At that point, even if a piece of the fuselage had blown off, there’s no reason they couldn’t have deactivated autoland and taxied off the runway. How do they justify that action? Those pilots are suspect IMHO.
FAR 91.3 is one thing but so is good common sense, too. If I were the Manager of the FSDO with authority over the situation, the registered letter requesting a full explanation to include why they did that would have already gone out. It’s time for a NTSB investigation and — potentially — a new FAR IMHO. And if I were a Chief Pilot of any organization flying such an airplane, I’d be rewriting the rules of engagement, as well. Finally, I think Garmin should be rewriting their manuals, too. Engaging autoland put a whole bunch of people potentially at increased risk here. That’s a serious situation. What’s next, engaging autoland because the potty stopped working? What if this happened near NYC … think of how serious that would have gotten.
The best way to figure out what to do with one of those “magic buttons” in your airplane would be to first ask yourself … “WWCYD” … ‘What Would Chuck Yeager Do’ before pressing it. We don’t need wanna be Walter Mitty test pilots thinking it’d be fun to “try it for real” unless it IS real … REAL real!
Be careful what you recommend, here … you just might get it.
I didn’t watch the video until after I typed the above comment. Now having done that — in a way — this situation gets worse.
First off, the installing shop can’t fully test the system after installation and ‘isn’t allowed to’ per Blackhawk. That in itself shows the seriousness of using autoland. Will it work or won’t it; it’s a crap shoot. Full autoland checkout doesn’t happen until you use it. I don’t like that idea. And using the system assumes that the airplane itself is 100% operational. What if it isn’t? Garmin can’t design a software response to an infinite number of situational / airframe ‘what if’s.’ Autoland should be the mode of absolute last resort and ‘we’ve tried everything else and give up’ sort of thing. IMHO, a RD / EDM activation carried into a full autoland mode isn’t one of ’em.
Next up, it’s not clear if this King Air went into ‘Emergency Descent Mode’ (EDM) by itself or it was commanded into a full autoland situation by the crew right from the beginning. Assuming a depressurization / EDM above 20K’ triggered the system, the crew would have had to communicate with the system to continue to a full autoland mode once it reached a safe altitude. I’d want to know IF that happened and what were they thinking when they commanded that, IF they did? Why didn’t they just take over?
I pondered whether it was possible to turn autoland off. Now I know you can. So however this situation started — automatically or commanded by the crew — it would have been possible for them to disengage and hand fly the airplane once it was at a safe altitude and away from the mountains.
I wondered whether crew or pax could still use the radio when the system was activated. From the screens in the video, it appears that would be possible. So now I’d want to know why qualified pilots weren’t communicating? What happened to “Climb, Confess and Comply?” Had the airplane crashed and without CVR or flight recorder aboard, investigators would have very little to go by to figure out what happened.
Finally, I saw a screen which warns the occupants to not deplane until the props aren’t turning. I saw them nonchalantly exiting carrying their things with the props still turning. Real professional, guys 🫤
I said it the first time and now I’m ever more convinced, this was a poorly trained crew who overreacted to what was a pan, pan, pan situation. It’s nice that all’s well that ends well but … there’s always more to the situation and room for improvement.
Larry, I agree on a flight test before returning the airplane to service, but that last step, as it stands, is not on the shop, it is on the owner or operator. The avionics guys are boxed in by the STC, the FAA, and the insurance underwriter; they cannot just go fly a full system check because Emergency Autoland is set up to run as an actual emergency and, as far as I have seen, there is no dedicated non emergency flight test mode.
That not only limits what the shop can verify, it also makes it harder to train pilots on the full system without turning every exercise into a simulated emergency.
Garmin really ought to incorporate a flight test mode that does not turn into a declared emergency.
I didn’t mean to imply that the shop should — or could — fully test the system, Raf. And I fully agree that Garmin and the individual operators had better revisit the certification and training process. There are now 1,700 autoland equipped airplanes in service and growing.
That said, I stand by my position that these two pilots took this situation FAR beyond where it should have logically been terminated. Once they got the situation normalized and autoland wasn’t having any problem flying the airplane, why couldn’t the pilots terminate autoland, communicate the situation and fly the airplane to the destination? Further, as the Blackhawk employee said, once the airplane descended to the design ‘safe altitude’ after they RD, THEY had to additionally command the system to autoland. Sorry … that ain’t a logical decision. They shoulda pushed the autopilot disengage button and hand flown it; that’s what I’da done.
I found the following on Flight Global this morning:
“Due to the complexity of the specific situation – including instrument meteorological conditions, mountainous terrain, active icing conditions, unknown reasons for loss of pressure and the binary (all-or-nothing) function of the Garmin emergency systems – the pilots, exercising conservative judgement… made the decision to leave the system engaged while monitoring its performance,” Buffalo (the operator) says.
The pilots were prepared to take over if required, but the system worked “exactly as expected”, it adds.
The FAA is investigating. It provides few details other than to say the “pilot lost communication with air traffic control”.
HUH! ? Autoland didn’t have any problem communicating … but the crew couldn’t??
The FAA better jump on this situation pronto. A change in the design and certification of the system — as you opined — a change in crew training and an additional FAR covering use of it are also called for IMHO. I’m a FIRM believer in the axiom that, “IF it’s working, don’t mess with it” but … this situation reveals that more work needs to be done to incorporate autoland into the NAS. Their continuation of the use of autoland put others in the system at potential risk.
The only thing that’d have made this situation even nuttier woulda been if the crew ordered up a Uber car and it was waiting for them on the runway where they left the airplane spooling down as they sped away.
I hope the FAA and Garmin are paying attention here … ??
The more I think about it, these two pilots must have been former Airbus drivers ?? Boeing jocks woulda flown it home by hand.
WWCYD !!
Anyone know how much an auto land installation costs, like in a King Air?
Good point, Roger. Based on the video interview. For that early 1980s King Air 200 with a full G1000 NXi “Platinum” suite, new RA/SurfaceWatch, plus Autothrottle and Autoland, and about 300 man-hours of install time, figure, guesstimate, roughly half a million dollars.