Autonomous airplanes, Anonymous Pilots

ArticlesTraining

Posted on Apr 4, 2020 by Trip Taylor

I am curious how one could program an Airbus to fly autonomously to a landing in the Hudson river.  As advancements in technology exceed expectations almost daily, the pause between the next new gee whiz announcement and the previous one leaves precious little time to contemplate whether the last amazing feat was, in fact, a wonderful feat.

In our Cirrus airplanes, we are treated to among the highest level of autonomous flight available in aviation.  In a new Perspective after take off, the YD can come on to alleviate the need to use our right leg, the AP can be engaged at 500’ and the pilot need not operate the controls until the airplane is 200’ from landing.  If we go too slow or bank too far, the plane will nudge us toward a sustainable attitude.  My client’s brand new 2016 Cirrus SR22 got a 25 hour workout last week.  The key fob unlocked the doors, the air conditioning kept us cool and the Perspective system allowed us to, almost autonomously, fly dozens of procedures. 

Learning all of the ins and outs of a comprehensively integrated airplane presents a challenge for both instructor and student.  In the past, instrument training included learning to fly the airplane by hand via an assortment of flight instruments that are common to many airplanes.  Avionics in such airplanes are fairly standard.  If you have worked one VOR receiver, you can probably work any VOR receiver.  However, as the fleet of Cirrus airplanes evolves, even the factory flight department finds it harder and harder to support previous generations of flight decks.  Fortunately, the vast experience within COPA and the training teams of both organizations work together to support the legacy fleet in the training realm. 

Many Cirrus pilots come to the airframe with good basic skills and most have at least a private certificate.  However, even for the instrument rated pilots, transition into the Cirrus to a level of safe operation in actual IMC conditions takes a significant commitment.   My colleagues and I spend a great deal of time comparing notes with the goal of advancing the state of the state of the art of Cirrus procedure training.  Art seems an appropriate word with respect to this challenge. 

The expected outcome of the training is for the new pilot to be able to operate the automation fully, to be able to operate without the automation, to discern when the automation is operating incorrectly and to be adept at diagnosing the malfunctions.  It’s a tall order.  As we wrestle with those dragons we are incessantly reminded of the high profile Air France Airbus accident and other instances where the command pilot appears to have failed to operate the airplane at a basic airmanship level.  Thus, as educators, we have to be very careful in changing the manner in which we train pilots of complex integrated airplanes. 

One of the core principles of instruction is the demonstrate/preform model.  The instructor demonstrates how to execute the maneuver, explains said maneuver and the student then attempts the maneuver. 

Another staple of Instrument training is Basic Attitude Instrument Flying (BAIF).  Instructors are well acquainted with the resulting Barely Able to Really Fly (BARF) gyrations as the student learns the nuances of the instrument scan.  For me, the BAIF period came first and, slowly, additional tasks were added.  At issue is the reality that very little brain is left to learn how to master a complicated FMS system while wrestling our previously tame airplane into submission. 

Enter the structured approach.  The goal is to divide the new learning from the new motor skills.   Foundationally, the same skills are taught but we flip the old model upside down. 

In my practice, I break this down into 3 phases.

  1.  Learn how to make the airplane perform and master the configurations, attitudes and power settings. 
  2. Learn to operate the airplane in a fully coupled manner
  3. Start taking away parts of the magic until the pilot is capable of hand flying all procedures both via partial panel and while programing the flight director (if so equipped).

This model is the reverse of how I learned.  In working with dozens of clients, I have had good results as we have followed this plan.

A recent example highlights the challenge of learning a complicated airplane and making it do a simple task.

The VOR RWY 33 approach into Chattanooga, TN (KCHA), is a fairly straightforward affair.  (Chart reproduced for your pleasure).  A local examiner always includes this approach as a circle to land, non-precision approach for the Instrument check ride.  I have instructed students on how to fly this approach hundreds of times.  I had never flown it myself.  I am sure any of my students that read this will be eager to give me an emphatic, “See, I told you it wasn’t so easy despite what you said” ribbing.  Yet, I truly struggled to demonstrate this approach to a client early in his instrument rating training just the other day. 

In the old Bonanza that I flew a while back, this approach would be simple, tune and ID the station, Set the inbound course and receive vectors from ATC.  When the To/From flag flips descend to MDA and circle to land. 

In the Perspective, I needed to verify the radio ID and make sure the automation changed the source.  If using the flight director, I needed to see a VAPP display on the top of the PFD and I had to verify station passage to begin the descent.  And for that descent, did I want Flight Director Guidance?  Well, I would have to program that too.  We determined it wasn’t worth the effort and simply flew the needles outbound from the station passage to the MDA. 

As you no doubt remember, the short distance just before and after VOR station passage is referred to as the “zone of confusion”.  And apt moniker if there ever was one.  Compared to the stately descent on glide path or glide slope on so many approaches, this VOR approach feels frenetic and unsettling.  Therein lies the crux of our training dilemma.  We must use the automation as a tool to reduce our workload rather than in impediment to an orderly flow.  It became clear to me that approaches using old school navaids may best be taught without the aid or complexity of added automation.

Within the Cirrus Pilot Proficiency Program (CPPP) we have introduced a concept of the Six Sacred Skills.  And some debate has ensued about the value of the skills.  We believe that our fully integrated airplanes lack one layer of integration.  We need to integrate the pilot to a new awareness of how the on board equipment should be able to give the pilot the freedom to take over when things go pear shaped.  In order to integrate the pilot to the flight deck, we feel that digging a bit deeper into the capability of the avionics can produce a pilot that is able to discern when a higher or lower level of automation is called for.

Because, you see, the best avionics package in the world cannot land the airplane safely on the Hudson river.  Only a pilot that has studied the airplane thoroughly is prepared to make that split second decision to chuck all the automation out the window and fly the airplane to the safest outcome possible.

Theme picker

Join COPA today and start saving!

You might also like

ArticlesTrainingSafety

Flying in Snow

Nov 26, 2023
ArticlesTrainingSafety

Icing

Nov 26, 2023
ArticlesPartner Perspective

Becoming a ‘Crew Resource’ in Flight

Aug 1, 2023