Computer culture clash?

April 1, 1998

Very few people, if any, are killed by computers while sitting at their desks, although more than a few computers have suffered mortal wounds at the hands of frustrated humans. Computers have killed in aviation, however — and while the pilot always shares some culpability, the failure to communicate between the two is much more serious than a document deleted from the desktop. Computers will play a much larger role in aviation, and the big question is whether they will make navigation easier or more complex.

The most critical computer cockpit function will come when GPS becomes the primary means of airborne navigation. Compared to VOR/ILS, present GPS receivers require two to four times as much programming in order to perform similar tasks.

GPS has brought about dramatic changes in the way that we navigate and provides tremendous new capabilities to pilots, particularly those who have already embraced IFR GPS technology. The price we have paid is the requirement that we learn different operating systems for every different brand of unit, some of which demand more than a little button-pushing to get from A to B.

The FAA is about to set the standards for a new generation of units, the ones we'll be living with for a long time. The AOPA Air Safety Foundation spotlighted the need for simpler user interfaces last year by hosting a conference with the FAA and the manufacturers, where a dialogue on some basic standards was begun. The General Aviation Manufacturers Association agreed to take on the facilitation role so that the ASF could function as a pilot advocate.

The ASF is asking for a degree of commonality in only a few core functions. This includes tracking direct to and from a waypoint, selecting an instrument approach procedure, initiating a missed approach, and entering a holding pattern. Everything else is open for innovation. However, common labeling of knobs and buttons — and commonality of the core functions that those buttons control — will make the transition to IFR GPS much easier. The manufacturers understand the need for simplicity, but most remain convinced that their particular way to navigate is the best.

The ASF conducted a random survey in February to assess the feelings of instrument-rated AOPA members who will be tasked with learning the operations of these new receivers. Several key questions were asked: Should terminology between receivers be standardized — in other words, should the knobs and buttons on GPS receivers be labeled the same from GPS manufacturer to manufacturer? Seventy-one percent said "yes."

Should the core functions mentioned above be standardized among receivers? More than 87 percent said "yes."

It was important that pilots understand what they might be giving up under standardization, so we asked the question in a different way. "In your opinion, what is more important — standardizing terminology and functionality of GPS receivers or allowing GPS manufacturers the freedom to design GPS receivers in their own manner?" By a ratio of two to one this sample preferred standardization in core functions over absolute freedom of design.

Looking at the different cultures of aviation and computers, it is easy to see why there is a disconnect. In the computer world, things change very rapidly; human factors testing is not usually a strength; and if there's a glitch in the program, no problem — you just buy the next version or download the upgrade off the Internet. Because of the complexity of the functions, it's unusual to get it right the first time. Standards are slow to evolve because there is so much innovation in the marketplace.

Within the aviation world, things change relatively slowly; human factors testing is not usually a strength; and if there's a glitch in the product, my survivors' lawyers will be in touch — you can download the subpoena off the Internet. Problems in the cockpit tend to be much uglier than on the desktop, and the aftermath can last an eternity. Because of the extreme consequences of mistakes, much testing is done. Standards, generally, have evolved early because of heavy regulation, which stymies innovation.

For most flight-critical functions, there are some standards: the "standard T" for flight instruments; throttles on the left, props in the middle, mixtures on the right; landing gear on the left of the quadrant and flaps on the right, etc. Where systems are not standardized, there tends to be difficulty. Beech used a different flap/gear configuration for years on the Baron and Bonanza series and both aircraft have significantly higher involvement in gear-up accidents. (One Beech friend pointed out that the rest of the industry chose wrong.) Pilots also hang themselves regularly by selecting the wrong tank or the Off position, making fuel systems another problem area.

True, present avionics systems are not standardized. However, flight-critical functions involving communication or VOR navigation generally can be sorted out with a short period of study. I didn't say that you could make full use of all the equipment or some of its work-saving features. However, almost anyone can get into a light aircraft and figure out how to get to point B.

Ultimately GA must evolve, but the computer engineers should remember that our work stations move at 100 knots or faster and pilots are unable, in many situations, to devote 15 or 20 seconds to programming. We've talked to owners who really liked their GPS units, but they flew only one aircraft and had spent hours and hours learning that receiver's nuances. We haven't heard from any renter pilots or CFIs who felt comfortable with having to learn multiple operating systems for a flight-critical component. No longer will GPS be the province of the early adopters, the technically proficient, or the gadget freaks. Everyone will have to learn, and increased complexity is not the way to bring new people into the aviation business. Simpler is safer.

Some training is essential. After all, you didn't get to be competent without some effort on VOR and ILS, and it is unrealistic to adopt a completely new way of doing business without some level of training. Likewise, it is unrealistic and dangerous to expect pilots to learn diverse operating logic and maintain proficiency on each unit. It's like having to learn several word processing programs. Historically, when new technology is introduced, there is a learning curve for both designers and pilots. A primary nav system is an excellent place to remember the human factors lessons of the past and to not rely exclusively on training to save the day.

Conversely, we don't want to stifle innovation and product improvements during the early stages of evolution by insisting on rigid standardization. The most compelling argument for diversity is that if standards are set too soon, we might just standardize on a DOS-type operating system. By allowing innovation, Windows 2000 could be just around the corner. That may occur in two years or 10. Note that this did not happen with loran — there was no standard, and IFR GPS is more critical because it will involve approaches. But since "computer think" caused the problem, it may also provide the solution, as DOS and Windows can both run on the same machine. A very simple default system common to all units would allow any pilot to perform core operations in a standard way and solve everyone's problem.

The natural resolution will occur when one of the manufacturers comes up with a killer application. That's computer jargon for a marketplace smash that is so simple and intuitive that everyone will be forced to emulate it or go out of business. In aviation, we prefer not to describe success exactly in those terms; but, paradoxically, a killer application will ultimately save lives. We urge the manufacturers and the FAA onto a path of intelligent commonality and simplicity — that's a concept all pilots and their passengers can live with.


The ASF would like to hear from you pro or con on GPS standards. Send e-mail to asf@aopa.org or write to the AOPA Air Safety Foundation, GPS Standards, 421 Aviation Way, Frederick, Maryland 21701.


See also the index of "Safety Pilot" articles, organized by subject. Bruce Landsberg is executive director of the AOPA Air Safety Foundation.