Human nature vs. hardware

August 1, 2006

Bruce Landsberg is the executive director of the AOPA Air Safety Foundation.

I'm always amused when "inquiring minds" want to know why the industry, the FAA, or somebody hasn't done something about the ongoing issue that pilot error consistently accounts for 75 percent of all accidents. As I've stated in the past, there are some rather compelling reasons why this 3-1 ratio doesn't yield easily to major initiatives. Fixing underperforming hardware is much easier to address than underperforming humans — we'll get to that momentarily.

The less discerning pundits both inside and outside of the aviation world constantly try to compare light general aviation with airlines using meaningless statistics such as, "The fatality rate of GA aircraft is 100 times that of the air carriers." There's no need to bore you with the obvious differences here other than to say that this level of comparative "wisdom" isn't often applied to ocean liners and small boats, buses and motorcycles, or the NFL and sandlot football. For some reason everyone seems to understand that large ships can negotiate heavy seas that would be foolish to attempt in small craft, buses by size and structure are inherently safer than two wheelers, and the sandlot players are simply incomparable to NFL behemoths. Regulators, insurers, and participants in the aforementioned activities understand the natural differences between the machinery and the oversight systems. They don't try to "fix" the unfixable. There are occasional fatalities but individual failures, as unfortunate as they may be, are accepted.

That doesn't mean GA can't do better. We can, but reasonable expectations and how we can meet those expectations are great topics for discussion. Education and training can often be used to good effect. This is AOPA's and your Air Safety Foundation's preferred method to help pilots understand high-risk areas and address them. Free seminars and online courses, publications, statistical analyses to present the facts, and free newsletters to all CFIs and flight instructor refresher clinics are just some of the products that are there to help those who want to be helped. The medicine is effective for those who choose to take it, but some say we're preaching to the choir.

The 'Bozo' factor

For the airlines and many corporate pilots, good training, mentoring, and an organized system of rewards/punishment play a part. Their jobs are on the line and there is a deep layer of oversight — something that would be logistically complex and very expensive for GA to implement. But education alone is only partially effective on the few who cause a disproportionate number of GA accidents, in my opinion. It's impossible to measure the "Bozo" factor, but my sense is that most of GA's bad actors are a small percentage of the flying population. Positively identifying Bozos in advance of a mishap is devilishly tough. Problems arise in three areas: ignorance, carelessness, and arrogance. We probably can't do much about the last group, but there are some things that could work for the first two.

Human nature is difficult, perhaps impossible, to change. We forget, we are distracted, we get fatigued, we get overwhelmed, and then there is the ever-popular brain fade. Any or all of these could be present during a flight. There are some incremental gains that can be made from training, but it's so much easier to design the technology to fit the human in the first place. Instead, we push, pull, threaten, plead, bluster, and drill to force pilots to learn the machinery. This process must be repeated for every new pilot, and refreshers are essential for the experienced men and women whose humanity and memory also are liabilities. Unfortunately, the human nature of some engineers and manufacturers is to overlook the human nature of their customers.

Now, how about hardware? The most consistent major safety improvements in air-carrier aviation have come as a result of technology. Jet engines have done more to improve safety than anything else: Reliability and lots of power go a very long way to eliminate failures in critical flight phases and to get over much of the weather that plagues those of us using pistons. Good instrumentation, intelligent automation, and redundancy also help a lot. Here are just two areas where some clever engineering will outdo training and pilot punishment every time.

This is so basic that most pilots' eyes glaze when you discuss it and yet in the two following accident categories, better technology could save millions of dollars and a few lives every year. Big accident producer: fuel mismanagement. Over the past several years we've been cracking up nearly three aircraft a week because of fuel mismanagement. I've previously commended Cessna for what seems to be a nearly bulletproof warning system on its new piston singles. It has a separate monitor on each tank and when the fuel gets below 30 to 45 minutes in that tank, an annunciator flashes and cannot be reset. At last report, with more than 5,000 units in the field, there have been no fuel-mismanagement issues — gold star!

Other manufacturers are using similar technology successfully, and we can only hope that as more new aircraft come into the fleet the appalling fuel-management problem will begin to abate. This is simple to build in and expensive to retrofit. The good news is that no additional training is required.


The other low-hanging fruit is landing-gear mishaps. There are about as many gear-ups as fuel fiascos, but many incidents never hit the NTSB's database since they don't rise to the requirements of NTSB Part 830. Check out in a retractable and you will always hear the tired joke of the two kinds of pilots who fly retractables. Some manufacturers have solved the problem by eliminating the folding wheels while still giving high performance. A lot of engineering went into that solution — a new type certificate. Elegant answer but retractable gear will always be with us in some aircraft.

The training/procedural approach has been attempted since the first belly flop but as we apparently have yet to learn, a hard-wired human weakness is the inability to function well when distracted. The accident/incident reports are rife with pilots ignoring horns, lights, bells, and other micro-switch-activated warnings that attempt to save us from ourselves. It's only partially effective.

Piper Aircraft, some 30 years ago, came up with a highly effective automatic gear lowering system that made pilots really work to belly flop. It had a pitot-tubelike air-pressure sensor that lowered the gear when the airspeed was below about 90 knots and flashed a light as well as sounded a horn to remind the pilot that the system had just saved him from a major embarrassment and expense. Beech also had a similar system, called "Magic Hand," on some early Bonanzas.

The system wasn't perfect as some enterprising attorneys soon discovered. If the sensing mast was blocked by ice or the aircraft suffered a loss of power and airspeed went below the magic number, the gear would extend when you really didn't need it. Never mind that the pilot's operating handbook required the pilot to merely engage an override lockout switch that commanded the gear back up and made everything operate in a conventional manner. The system was successfully sued out of the design, and a service bulletin recommending that it be disabled was issued. While the regulatory and FAA focus was on the negative, little attention was paid to the benefit of thousands of prevented gear-up landings. Something about babies and bath water comes to mind.

Now jump ahead to the second century of flight, and see that the problem is still with us. Here's a modest proposal: How about using the GPS to act as a safety pilot? When the flight is within a mile or so of the airport and the altitude is below, say, 500 feet, if the pilot hasn't put the gear down, the system does it for him. It then advises the pilot and sends a datalink message to his insurance company and CFI — just kidding about the second part. As a guard against improper deployment the system also would require the power setting to be less than 50 percent. There is probably something I'm forgetting here, but you get the idea.

The system

The conventional thinking is that pilots would become dependent on it and the system could fail. Both true, but I'll take those odds for two reasons. First, most pilots take it as a point of honor to stay ahead of their machines. In my nearly 500 hours of flight time in "protected" Piper Arrows I never observed any pilot taking pride in the fact that the system had just saved his bacon. We never had a system malfunction either. In many boardrooms and within some areas of the FAA's certification process, way too much attention is placed on the 100-percent solution when 98 percent of the time the system will work and save us all a lot of grief and expense. Getting the last 1.999 percent is prohibitively complex and expensive. So a lot of good ideas that would make the piloting of aircraft easier, safer, and less expensive don't make it into production.

Maybe someone has already come up with this and we'll see it on the next round of retractables. What good ideas do you have that could make our machines more friendly? Let us know what you think.