An interesting article about the downside of technology/automation

I actually think the automation isn't the issue, but rather it allows the pilot to be less engaged.  There is a basic rule of flying any size or shape airplane, "first you fly the plane"!  Getting all caught up in the automation is an accident waiting to happen. 

FWIW, this is not unique to the 737 MAX, there are a number of accident reports from the NTSB over the years where pilots got so dependent on the automation that they really didn't have the skills to fly the plane.

Actually, the first rule of flying for pilots is, "Landings = Takeoffs" 

Well, we could start spouting the old saw "Any landing you walk away from is a good landing", but I prefer to be able to fly the plane again after the landing.

I confess, the automation is somewhat addictive.  I flown several coast to coast trips, and many more up and down the East Coast.  Having a 3-axis autopilot made the trips much easier, just set and you're basically a passenger for the next couple of hours.  However, not being constant vigilant is a recipe for disaster, it's not like being in a car when you can just pull over if something goes wrong!  You still have to manage the flight and know where you are and the status of the airplane.  Also, if you're VFR, you are responsible to see and avoid other traffic.  I once had a vacuum pump seize over some pretty desolate territory out west, and of course the autopilot immediately stopped flying the plane.  I was suddenly in a fairly steep climb, SURPRISE!   Think about what happens in a similar situation if you're depending on the autopilot to fly an approach and it decides to drop off and points the nose down!!!  If you're close enough to the ground, before you can react, you have that "landing" you speak of, only it's not the smooth affair you were planning on!

GRJ, yep, automation can be addictive and I appreciate a pilot's perspective. One of the biggest problems with any kind of automation is so called human factors engineering, and a big one is how do you keep people fully engaged when you have a system doing so much?  This happened in the securities industry when they started heavily using algorithmic based trading systems, those monitoring the activity being done even though there were tools written and reports generated, were found over time to become lulled with a false sense of security, that everything was fine, rather than applying diligence beyond what the reports/consoles were showing them, and some companies ended up with major problems when systems went south when market forces caused them to thrash. 

Doesn't mean automation is a bad thing, just that often it is introduced without thinking about the side effects and consequences and in making it work seamlessly with human operators, how to keep them engaged. I saw a video on you tube the other day, this blog by a guy who is a long haul trucker, and he was showing all the tech he had in the cab of his truck (not gonna mention the sleeper, wow), that allowed all these different things they could monitor, etc, and I wondered if driving down the road with all those distractions, how easy it would be to get lulled into looking at them rather than the road, kind of a similar thing. 

A friend of mine got a new car about a year ago for his wife, and he got all the new automated safety devices for lane changing and braking and cruise control.     They drove it to Columbus (we live in Mich) for Christmas and it was a snowy, slushy misty day which we have a lot of in the north.    Anyway he realized the cruise control stopped working and then other stuff.    He thought it was broken.   But when he stopped and cleaned off the camera stuff, it would start working again for about 5 minutes.    So all this automation which I assume works well in sunny california or someplace was a complete failure when he needed it most on a snowy day in Michigan and Ohio.     these automation gurus need to figure out how make the stuff work everywhere not just under ideal conditions.    

In my years of aerospace development, we tended to try to make the stuff work in all weather, you never know what climate your airline is flying to.

I guess if it's snowy enough for the camera to be obscured, you probably shouldn't be using stuff like cruise control or lane guidance.   I'm pretty sure the anti-lock brakes and traction control was still probably working.

And then there are the cautionary tales about AI.......

Tinplate Art posted:

And then there are the cautionary tales about AI.......

Skynet woke up and within hours declared War on Humankind.

Lew

I am an active pilot who flies a Beechcraft King Air 350 almost every week, on 600 to 1,000 mile trips from NE Ohio to places like Key West, Oklahoma City and many other places where the boss needs to go.

I have seen first-hand how basic stick and rudder skills in aviation are deteriorating with the progression of automated systems in the cockpit.  We always fly the King Air with a co-pilot, even though it is not legally required. I have flown with a lot of guys in the right seat who would be totally useless in any kind of emergency or if the magenta course line on the Garmin G1000 suddenly disappeared. One only needs to watch a few episodes of “Air Disasters” to see how this has affected even experienced airline pilots. Landing short of the runway at San Francisco, or riding a stalled aircraft from FL 330 (33,000 feet) all the way to a crash in the ocean are two PERFECT examples of highly paid commercial pilots who simply did not know how to fly an aircraft.

The King Air has an auto-pilot, and I use it a lot in cruise. But “Otto” and I have an agreement when it comes to flying instrument approaches. We “share“ those duties. He flies one and I fly one. That way I know that Otto can still fly an IFR approach and it makes sure I can still fly one, too.

We already have a generation of young engineers who don’t understand how to properly use the air brakes, don’t understand how to control slack and really have no understanding of train dynamics because they haven’t been taught about those things.  They have only been taught what they need to know to get a train over the road with everything working properly. Let something go wrong and they are totally lost and unable to do any creative problem solving because they don’t even know what the problem is, let alone how to solve it.

Automation is not always a good thing.

FORMER OGR CEO - RETIRED posted:

 

We already have a generation of young engineers who don’t understand how to properly use the air brakes, don’t understand how to control slack and really have no understanding of train dynamics because they haven’t been taught about those things.  They have only been taught what they need to know to get a train over the road with everything working properly. Let something go wrong and they are totally lost and unable to do any creative problem solving because they don’t even know what the problem is, let alone how to solve it.

LOL, that kinda' reminds me of a humorous commercial that used to be on TV, where (if memory serves) a business-type guy carrying a briefcase and talking on his cellphone is taking an escalator to go upstairs, and all of a sudden, the electricity goes out.  He's stranded halfway between floors and doesn't know what to do!

That's modern technology for ya'.  

FORMER OGR CEO - RETIRED posted:

One only needs to watch a few episodes of “Air Disasters” to see how this has affected even experienced airline pilots. Landing short of the runway at San Francisco, or riding a stalled aircraft from FL 330 (33,000 feet) all the way to a crash in the ocean are two PERFECT examples of highly paid commercial pilots who simply did not know how to fly an aircraft.

The King Air has an auto-pilot, and I use it a lot in cruise. But “Otto” and I have an agreement when it comes to flying instrument approaches. We “share“ those duties. He flies one and I fly one. That way I know that Otto can still fly an IFR approach and it makes sure I can still fly one, too.

We already have a generation of young engineers who don’t understand how to properly use the air brakes, don’t understand how to control slack and really have no understanding of train dynamics because they haven’t been taught about those things.  They have only been taught what they need to know to get a train over the road with everything working properly. Let something go wrong and they are totally lost and unable to do any creative problem solving because they don’t even know what the problem is, let alone how to solve it.

Automation is not always a good thing.

I'm fascinated by some of the episodes of Air Disasters, it boggles the mind how some of these crashes happen!

FORMER OGR CEO - RETIRED posted:

The King Air has an auto-pilot, and I use it a lot in cruise. But “Otto” and I have an agreement when it comes to flying instrument approaches. We “share“ those duties. He flies one and I fly one. That way I know that Otto can still fly an IFR approach and it makes sure I can still fly one, too.

Good idea, that is smart flying. Too bad it's not more widely practiced. 

FORMER OGR CEO - RETIRED posted:

We already have a generation of young engineers who don’t understand how to properly use the air brakes, don’t understand how to control slack and really have no understanding of train dynamics because they haven’t been taught about those things.  They have only been taught what they need to know to get a train over the road with everything working properly. Let something go wrong and they are totally lost and unable to do any creative problem solving because they don’t even know what the problem is, let alone how to solve it.

Automation is not always a good thing.

A bit worrisome, perhaps why there's so many train accidents nowadays.

FORMER OGR CEO - RETIRED posted:

Automation is not always a good thing.

Yep!

RICH: Your very wise comments, based on solid experience, ARE indeed a "cautionary tale"! YIKES!

Between the Boeing fiasco, and the increasing over dependence on avionic automation, I now have a very different view of flight safety, to say the least! Again, YIKES! BTW, RICH, I would fly with you anywhere!

"Otto?"  His/her name is "George."

Airbus guy here.  I loved the systems, but knew how to turn them off.  I never flew a revenue flight in "direct law", but practiced ILS approaches on one engine with all control computers off.

And like Rich, it would be at least one "hand-flown" leg and one fully automated leg for each sequence of flights.  Unlike Rich, my copilots were usually as good at it as was I, and often better!

So here's my opinion on the latest kerfuffle.  Keep in mind that I am not reckless or even "brave" - I rarely get outside a very conservative box in aviation.

The Max 8 may have a glitch somewhere.  The obvious cure is to turn off the offending system and fly it by hand. (Well, the obvious temporary cure - don't let the airplane fly you).  Look at how many successful Max 8 flights have been accomplished.

Also look at the training of the crews on the two airplanes that crashed.  I heard that one F/O had 200 hours total time!  If true, unconscionable!  You will not find pilots flying US registered airliners with less than 1500 hours, and most have ten times that.  Experience matters.

If you are flying an automated aircraft, never simply sit there and ask "why is it doing that?"  If you do, it will fly you into a mountain or worse.  Disconnect that sucker, and assume the duties of pilot in command.

When airliners and trucks go pilotless, I am going Amtrak.  When Amtrak goes pilotless, I am staying home.  A 200 hour copilot is darn near pilotless - these things need two highly qualified aviators.

Opinion, of course.  I would happily fly a Max today.

BOB2: GOOD points!

bob2 posted:

"Otto?"  His/her name is "George."

LOL! Like I have often said, "I don't care what you call me, just don't call me late for dinner!"

Unlike Rich, my copilots were usually as good at it as was I, and often better!

I'm glad you made this point because I didn't mean to disparage all of my co-pilots! Most of the guys I fly with are truly GOOD pilots with excellent stick and rudder skills. Many are better than I am because they make their living flying aircraft, whereas it has always been a sideline for me.  I've been fortunate to fly with a lot of guys who know what to role of the co-pilot truly is. For those of you who are not pilots, the role of the co-pilot is to make the pilot look good and take the ugly one at the bar!

Also look at the training of the crews on the two airplanes that crashed.  I heard that one F/O had 200 hours total time!  If true, unconscionable!  You will not find pilots flying US registered airliners with less than 1500 hours, and most have ten times that.  Experience matters.

A 200-hour pilot has not mastered ANYTHING yet. He doesn't even know what he doesn't know yet! If that is true, it is a stunning revelation.

A good friend of mine is a Check Airman for a major airline. For those of you who are not pilots, a "Check Airman" is the airline equivalent of a Road Foreman of Engines or a Designated Supervisor of Locomotive Engineers (DSLE) on a railroad. He is a supervisor of pilots and he does their check rides throughout the year.

My friend was involved in training the pilots of Asiana Airlines Flight 214 that crashed a Boeing 777 short of the runway in San Francisco. The cause of this crash was clearly determined to be pilot error. The pilot flying had not hand-flown a visual approach in years. Asiana had a policy which REQUIRED their pilots to engage the Auto-pilot just after they put the gear up on takeoff, and leave it engaged until just before landing. There is no way you can keep your basic stick and rudder flying skills sharp if you never use them! And in this case, the cause was that the pilot didn't even know how to properly use the auto-pilot in this unusual situation of flying a visual approach. He set the auto-pilot to the wrong mode (Flight Level Change) when on final approach. You would NEVER, EVER use that auto-pilot mode mode in this phase of flight! But because he was in an unusual situation of flying a visual approach, he didn't know how to do it.

Normally when training pilots in the simulator, training is done "to proficiency." This means that emergency procedures, flight maneuvers and IFR approaches are done multiple times until the pilot becomes "proficient" in his ability to perform the required maneuver or procedure to very tight performance standards. However, the training procedures mandated for the Asiana Pilots was different. My friend and the other Check Airmen involved in their training were only allowed to mark a maneuver or procedure as "Completed." The procedure may have been done very badly, but if it was within very broad performance limits, the training record was marked "Completed" and they moved on to the next item in the curriculum. My friend commented that he would NEVER get on an Asiana Airlines flight because the pilots the airline provided were some of the worst he had ever trained.

There is a parallel here in the railroad industry. New engineers are being taught the basics of WHAT to do, but they are not being taught WHY they should do it. If a new engineer has not been taught about the "WHY" of a given situation, he will not be able to recognize out-of-the-ordinary situations when they happen. He is helpless to know what to do to correct the situation because he won't even know that a "situation" exists until it's too late...just like the Asiana pilot.

As I said above, an engineer who has been trained like this can get over the road if everything is working perfectly. But let something different happen or something go wrong and we've got a recipe for disaster.