Career Magazine

Are Pilots Too Dependent on Automation?

By Aafo4ever @AAFO4Ever
Dealing with mechanical delays is one of the more unpleasant parts of my job.  I am a people pleaser (I've been told that's a bad thing) and explaining the complicated nature of a broken-down jetliner to a plane load of people wanting or needing to arrive at their destination is often difficult and unpleasant.  I know from experience as a passenger how frustrating it can be to sit in row 32 with a cockpit crew unwilling to make regular PA's, so I've always made a point to keep people informed...that isn't always good enough.
Are Pilots Too Dependent on Automation?
In the spirit of keeping the operation in motion, an aircraft used for commercial purposes is certified with a Minimum Equipment List...MEL for short.  Without an MEL, every single item on an airplane would have to be in working order for the aircraft to fly.  As you might imagine, even the simplest of airplanes are complex machines with countless moving parts.  Sometimes those parts don't work exactly as they should and that may or may not be okay.  Without an MEL, an airliner couldn't fly with a broken toilet seat, a coffee maker that wouldn't brew or a burned out reading light.  These items must be repaired eventually, but if they are listed in the MEL, then repairs can be deferred to a later date to facilitate an on-time departure.
Clearly there are items that cannot be deferred.  Engines, flight controls and safety items are required to be in working order and are not included in the MEL.  But the auto pilot, flight director and auto throttles are among the items listed in the MEL for the aircraft I fly and are the reason for this post.  The question is...are pilots too dependent on automation?  If we are, what happens when the automation fails?
Are Pilots Too Dependent on Automation?
The first jet arrived at the gate with several mechanical issues.  Two were minor discrepancies in the cabin that I expected to be remedied in short order, the third was more significant,  but I still didn't think it would delay our departure.  Then during the walk-around inspection, I found a fourth issue that had potential to ground the aircraft for at least a day.  Unfortunately, we had already started to board passengers by this point and, in the Captain's absence, I elected to let boarding continue.  Stopping the boarding process after it has already started is often more difficult than it sounds.  I would regret that decision later.
A mechanic arrived shortly and I hoped he would be able to shed light on the "go, no-go" decision that needed to be made.  But as we began to look deeper into the legal and safety aspects of the decision, it became clear that the Captain and I were not going to agree with the mechanic.  Of course as pilots, we have the right to refuse an aircraft, but I would much rather come to a mutual agreement.  There was some "grey area" involved, so our mechanic disappeared to confer with his superiors...that discussion would continue for well over an hour.  With restless passengers and a frustrated cabin crew, I made PA announcements to the passengers in 15 minute intervals until the decision was finally made to take the aircraft out of service for repair.  Everyone filed off the jet and waited in the lounge area until a new plan was announced.
The second jet arrived a few gates and a short walk away.  I started walking toward the new gate just before the announcement was made and many of the passengers, assuming I knew something they didn't, followed me through the terminal.  As I walked down the jetbridge onto the new jet, I could see a mechanic standing at the aircraft door.  Not a good sign.  The tower planner had assigned us a new jet that also had a mechanical issue.  I really wish they wouldn't do that.  Why on Earth they can't or won't check the status of an aircraft before assigning it to a flight is beyond me.  The second jet was out of service.  Thankfully, we knew this before we boarded the passengers.
We had to wait for jet number three.  It landed about 20 minutes later and parked at a gate about as far from where we were as possible.  The DFW airport is so big that it actually sits within the city limits of four suburban cities and we must have walked through at least two of them while traipsing through the terminal.  I arrived at the gate just as the inbound passengers started to deplane and elected to swim upstream through the river of exiting masses.  I wanted to check in with the crew to get a brief on the condition of the jet and get an early start on my pre-flight duties.  The news was not good.
Once again, we had been assigned a jet with a known mechanical discrepancy.  This was getting old...and embarrassing...but at least the passengers weren't on board yet and I wouldn't be responsible for PA announcements.  I was running out of things to say.
Apparently the autopilot had failed on the inbound flight.  A servo responsible for manipulating one of the ailerons was not working properly and would have to be replaced...a job that would take hours we didn't have.  I quickly referenced the Minimum Equipment List and confirmed what I already knew.  The jet could be dispatched without an autopilot.  There were some restrictions involved with this and the flight crew had to accept the condition of the jet, but we were legal to go.
Normally, I will not accept a jet without an autopilot on long flights, flights into poor weather conditions, flights into challenging or mountainous areas or on flights when I'm already feeling fatigued.  The automation provides a level of safety I'm not willing to do without in certain situations, but this time the flight we were about to embark on was less than two hours in length, the weather was good and, frankly, I couldn't imagine inconveniencing the passengers with another aircraft swap.  So off we went.
I learned a technique from my father, a retired airline pilot, who taught me to be proficient with all levels of automation.  He would fly the first leg of a trip completely by hand.  From lift off to touchdown, he would leave all automation off for the entire flight.  On the next leg, he would do just the opposite.  He would turn the autopilot on at the lowest legal altitude on takeoff and allow the aircraft to land itself at the destination.  On the third leg, he would use some combination of available automation...what ever he felt like using.  I must admit I haven't exactly followed his advice to the letter, but I do hand-fly the airplane far more than my peers.
We departed runway 18L and joined an RNAV SID (Standard Instrument Departure) right after takeoff.  The company encourages autopilot use for RNAV departures, but I typically hand fly them, so this was nothing new.  We headed west into the sun, climbing at 250 knots until passing 10,000 feet then accelerating to 310 and transitioning to Mach .76.  Climbing and descending without automation is relatively simple and doesn't take much more attention than normal.  It's the tedious nature of straight and level flight and the attention to detail required for an instrument approach that presents a challenge for many pilots.
We leveled off at Flight Level 320 (32,000 feet) and...well...the next hour or so was tiresome for me and would probably put you to sleep if I tried to put it to words.  The only excitement was almost constant movement of Flight Attendants and passengers in the cabin which required continual trimming of the flight controls.  It isn't something I notice when the autopilot is engaged, but a single passenger moving from seat to bathroom is enough to put the airplane completely off balance, so with constant movement in the cabin, trimming the jet was a full time job.
It seems silly and melodramatic to say now, but flying the entire leg by hand was mentally and physically exhausting.  I'm a perfectionist to the core, so 20 feet high or low just wasn't good enough.  I only blew the altitude once as I searched for an approach plate...60 feet high just before the top of descent.  I suppose that will have to be good enough.  By the time we landed, my eyes were weary and I felt generally fatigued.  We had one more leg to fly and I was pleased that it was the Captain's turn to fly.
The entire experience got me thinking about modern pilots and their reliance on automation.  I couldn't help thinking about pilots in the 30's and 40's flying DC-3s around the country in all kinds of weather with no autopilot, no autothrottles, and no flight director.  I'm tired just thinking about it.
The airplane I learned to fly in was equipped with a standard "six pack" instrument panel that was just barely legal for instrument flying.  It had one COM radio, one NAV radio, an ADF and basic instruments like you see in the picture below.  That was it...nothing fancy.  I got my tail wheel sign-off in a Piper Cub with no electrical system, no radio and only the minimum instrumentation required for visual flight.  I was taught to rely on on the basics.

Are Pilots Too Dependent on Automation?

"Six pack" instrument panel.

Conversely, I went flying recently in a beautiful A36 Bonanza owned by an old friend and college roommate.  Not only was his airplane better equipped than any general aviation airplane I've ever flown, but it was far better equipped than the MD80 I fly commercially.  As we prepared to start the engine, I was surprised that my friend didn't have any navigation charts handy.  We were departing underneath the DFW Class B and I wanted to know exactly where we were in relation to this restrictive airspace.  No need for maps, it was all depicted visually on his Garmin G1000.  Frequencies, airport diagram and approach pages?  No need for those was all available on the Garmin, but I couldn't help wondering what was going to happen if all his expensive electronics stopped working.

Are Pilots Too Dependent on Automation?

Garmin G1000 instrument panel installed in a Beechcraft Bonanza.

There is great value in getting back to the basics.  Even the airlines recognize that the continual reliance on automated systems is dulling the skills of their pilots.  When I landed my first airline job, the simulator training I received almost never involved the autopilot.  We hand flew every maneuver and only used the autopilot for straight and level flight.  As the years have gone by, I have noticed a continual increase in the level of automation either suggested or required during my training events.  
I clearly remember the first time I was allowed to utilize the autopilot for an approach during a training event.  I was an EMB-120 (Brasilia) Captain and we were training in a full motion simulator that emulated reality so well, I usually forgot we weren't in the the real airplane within minutes.  I couldn't believe they were allowing autopilot use during a training event, much less suggesting it.  The autopilot did a great job of flying the airplane, but it didn't say much about my skills.  I set it up, I pushed the buttons, I knew the legalities, and the airplane did a great job flying the approach...but could I have flown it if the autopilot had failed?  I knew the answer, but my instructor did not.
Normal for me is to hand fly the airplane to an altitude close to cruise, use the autopilot for straight and level flight and part of the descent, then turn off the automation for the approach and landing.  As a result, hand flying the simulator is not a big deal, but another issue forcing the increased use of automation by airline pilots is the growing number of RNAV arrivals and departures.   Most airlines encourage autopilot use during RNAV departures and arrivals due to strict course requirements and tiny deviations allowed on such procedures.  When my airline first started flying them, it was a requirement to use the autopilot on all RNAV procedures.  That's exactly when I noticed my flying skills beginning to dull.  I was constantly required to use the autopilot and wasn't getting enough practice with the basics.
As the RNAV procedures became more common and pilots became better and more comfortable with the flying them, my airline and others began to relax the strict requirement to use the autopilot.  This has helped my flying skills a lot, but as a direct result of the Colgan Air 3407 and Air Fance 447, my last trip through recurrent training included a hand flown ILS to minimums with no autopilot, flight director or autothrottle and training in the recognition and flying of an aircraft with iced over pitot tubes and static ports.  This was the first time in a very long time that I was actually required to hand fly an approach in the simulator.  Even as comfortable as I am hand flying the airplane, I was surprised how nervous I was with the idea of hand flying the airplane all the way to minimums.
Personally, I think it's about time.  Technology and automation in the cockpit is a significant factor in the continued decline in fatal aviation accidents around the world, but without basic skills to fall back on, pilots will be ill equipped in the event of a failure.  It's time to get back to the basics.

Back to Featured Articles on Logo Paperblog