Sunday, August 10, 2008

Complex Problems

The vast majority of pilots currently flying for the airlines received their training in a civilian setting, though that wasn't always the case. The usual route to the airlines is to earn your various certificates (either under part 61 or part 141), then decide whether to become a flight instructor or find another way to build the required hours to apply for a professional flying job. Some pilots become instructors and many of these folks provide quality instruction while building hours. Other pilots choose to build flight time through towing banners, flying traffic watch, performing pipeline patrol, or doing aerial survey work.

One thing that these U.S. pilots have in common is that they all earned a commercial pilot certificate and when they took that commercial check ride, it was in a complex aircraft, defined in 14 CFR 61 as an airplane that has a retractable landing gear, flaps, and a controllable pitch propeller or, in the case of a seaplane, flaps and a controllable pitch propeller. The problem is that finding a complex aircraft to train in is getting harder and harder. There are several reasons why this is the case.

There are only a handful of piston engine complex aircraft still being manufactured and the Piper Arrow the only complex trainer I know of that is still in production. Many large part 141 flight schools use Beech Bonanzas and some flight schools use twin-engine aircraft (more on this later). Otherwise, a potential commercial pilot candidate is left to choose from a selection of older Mooney, Cessna, and Beech aircraft that make up the aging fleet of complex trainer aircraft still in service.

One reason that few manufacturers want to produce a complex trainer is that many owners and operators know that maintenance and insurance costs are higher for retractable gear aircraft. And then there are those unintentional gear-up landings. New aircraft manufacturers like Diamond, Cirrus and Cessna (nee Columbia) rely on the simplicity and low operating costs of a fixed landing gear coupled with aerodynamically efficient airframe designs to sell aircraft. A fixed gear reduces the cost for periodic inspections, maintenance, and insurance, but it leaves aspiring commercial pilots, flight instructor candidates, and the instructors who train these pilots with fewer and fewer options.

Some flight schools and pilots get around this problem by earning their initial commercial or instructor certificate in a multi-engine aircraft, which allows them to do a single-engine add-on to that certificate in a fixed-gear aircraft. With the rising cost of fuel, this alternative is less attractive, but often the only option available. And guess what? The aging fleet of multi-engine trainers is in the same sorry state as the fleet of complex singles. Again, there is really only one light multi-engine trainer aircraft currently in production: The Piper Seminole.

One solution to this problem would be for the FAA to provide some relief by redefining complex aircraft for the purposes of training toward a Commercial Airplane Single-Engine certificate so that it includes certain fixed-gear aircraft with constant-speed propellers. This would vastly increase the number of available trainers, but I'm not sure anyone at the FAA is listening or willing. Without some sort of relief, the shortage of complex single and multi-engine aircraft will only worsen, leaving aspiring professional pilots out in the cold.

For another illustration of increasing complexity, you don't have to look any further than the G1000. I recently helped the owner of a G1000-equipped aircraft comply with a service bulletin that requires the testing of the Course and Baro knobs. Most everyone who does G1000 transition training wonders why the folks at Garmin designed the course and barometric pressure setting knobs so that they were concentric. These two functions, course and altimeter setting, have absolutely nothing to do with one another, so the reason must have been one of economics or the result of a committee decision.

Whatever the genesis, the design is really unfortunate and this particular service bulletin arose because some of these concentric knobs were failing in a subtle way: Setting the altimeter or adjusting the course in an affected G1000 unit could cause unwanted movement of the other setting. As my mom used to say "You've buttered your bread, now you can sleep in it."

If owning and maintaining a G1000-equipped aircraft weren't complicated enough, a recent instrument candidate I recommended was chided for not having memorized the model numbers of each and every component that makes up the G1000 system. This candidate was also told they should have known how each unit communicated with the others. My reaction was "Is this pilot supposed to demonstrate that they can fly a G1000-equipped aircraft or are they supposed to be able to fix it?" And where does this level of minutiae stop being useful? Is it sufficient to know that the units are connected with an ethernet network or should a candidate be able to describe the data packets that are exchanged, whether or not the packets are acknowledged, and what the bit ordering scheme is?

The FAA's flight instructor material describes four levels of learning, from lowest to highest:
  • Rote
  • Understanding
  • Application
  • Correlation

The various Practical Test Standards published by the FAA say that flight instructors are to train their candidates to the higher levels of knowledge - application or correlation. Memorizing the model numbers of the G1000 components is clearly the lowest level of knowledge, but it's also the easiest to test: You either know the letters and numbers or you don't.

Asking someone to explain how the various G1000 components play together would seem to be testing a higher level of knowledge, but on closer examination this is really of little use once you are airborne. When a G1000 component fails or misbehaves in flight, there is little the pilot can do other than maintain control of the aircraft, land, and let an avionics technician replace one or more of the pricey LRUs (line-replaceable units).

One thing Garmin could do to reduce complexity would be to redesign the G1000's MFD status display so that it would describe each LRU with a plain English name. Rather than demanding that a candidate know the model number of the G1000 air data computer, examiners could ask about real world scenarios, like what would your PFD look like if the air data computer failed.

If something isn't done to manage this complexity in an intelligent and useful way, pilots will be forced to learn more and more useless trivia and safety will be compromised. Some say that it already has.


Shdwcaster said...

Great article, although I think you forgot about Diamond's DA-42 Twinstar as a currently manufactured multi-engine trainer. In my local Seattle area, at least, it seems to be fairly popular for this purpose, even with all the engine issues with the Thielhart diesels.

John said...

Good point. I thought about mentioning the DA-42, but didn't for several reasons.

One reason is that, as you mention, Diamond is still working on a replacement for the Thielert engines since that company declared bankruptcy several months ago. It appears two options will eventually be available for the DA-42 - a different diesel engine design or a more conventional Lycoming IO-360.

Another reason for not mentioning the DA-42 is the cost of owning, operating, and renting these aircraft is significantly higher than a Seminole or an older Duchess. I wrote a while back about the price/performance aspects of the DA-42.

These higher costs make it hard for me to consider the DA-42 as a trainer, though in most other respects it appears well-suited to the task. I've never flown one, but have significant time in the DA20 and DA40 and have found those to be very well-made trainers.

asianbadger said...

"My reaction was "Is this pilot supposed to demonstrate that they can fly a G1000-equipped aircraft or are they supposed to be able to fix it?"

No kidding. I'm not real happy with the way the G1000 is being taught or tested.

As usual, a great article on the issues facing GA today. Thanks.

Eric said...

The DE that runs exams for my flight school has done a few G1000 instrument checkrides for my students, and he seems to take a very practical approach. He's a former airline pilot, and his checkrides in general are very real-world, so he mostly focuses on the students' ability to perform and not their rote book knowledge.

Then again, part of the checkride simplicity could be due to a lesser understanding of the G1000 systems.