Airplane manufacturers’ wholesale adoption of glass panels has been driven by market forces, but those layouts were designed to provide some safety advantages as well—chiefly improved situational awareness. The hope was that the very large attitude indicator in the primary flight display showing blue sky and brown earth would be harder to overlook or misinterpret, while detailed moving maps with terrain depiction would help prevent pilots from getting lost or prematurely descending below safe altitudes. (The goal of workload reduction has been more elusive; the increasing sophistication of modern autopilots is offset to some degree by the growing complexity of programming them and the nav sources they track, not to mention of locating the profusion of weather, airport, engine, and reference data lurking somewhere behind the panel. Building real proficiency requires considerably more practice than with earlier generations of GPS navigators.)
Any actual effects the glass revolution might have had on flight safety haven’t been easy to gauge, in part because it took a while to accumulate enough real-world experience with the equipment to support any evaluation. Time would have cured that—but the adoption of glass also coincided almost exactly with the certification of some new and very distinctive aircraft designs. Almost all the production of Cirrus and the certified Lancair line (renamed Columbia before it was bought by Cessna) has been delivered with glass cockpits as standard configuration, so a simple glass-to-analog comparison won’t work: You get most of the Skyhawks and Cherokees on one side and almost all the SR22s on the other. Clearly, these are very different groups of airplanes.
A new study by the Air Safety Institute has begun to untangle some of these complexities. It is restricted to certified piston airplanes manufactured since 1996, minimizing aging-aircraft issues unrelated to avionics design and defining a more stable segment of the fleet. Comparisons can be made both within and between groups of more-or-less comparable airplanes. In some cases, enough data exist to compare glass and analog directly within the same model lines.
When some of these confounding factors are pushed aside, we find that glass hasn’t changed things very much. The big differences are still between different classes of airplanes and the way they’re typically used. Low-powered fixed-gear singles (the Cessna 172s and Piper Cherokees) see a lot of duty as primary and instrument trainers and suffer a lot of low-speed, good-visibility accidents as a result. Their accident rates are high, more than double that of the faster and more powerful models, but fatal accidents are comparatively rare. Singles of 200 horsepower or more and light twins, on the other hand, spend far more time in cross-country personal and business travel and do more flying at night and in IMC. More experienced pilots plus less time in the pattern equals fewer takeoff and landing accidents, but going faster doesn’t help when you hit something you couldn’t see. While accidents in these models were only about half as common, they were three to four times more likely to be fatal, and fatal accident rates were higher as a result. Interesting differences also emerged between the new designs certified by Cirrus and Columbia and the long-established models offered by Cessna, Piper, Mooney, and Hawker-Beechcraft.
Within each of these groups, there was next to no difference between the records of glass-panel aircraft and those of similar models equipped with analog instruments. This may be less surprising when you note that more than two-thirds of all accidents still take place in daytime VMC, the conditions under which the improvements in situational awareness claimed by glass have the least practical significance. One finding, however, was both surprising and consistent across all the types of aircraft in the study: The glass-panel versions suffered more accidents during takeoffs, landings, and go-arounds, exactly the time you might expect panel configuration to be least relevant. The reasons still aren’t clear, but it’s fair to guess that most GA pilots don’t find the “tape” displays of airspeed and altitude easier to read than needles on a gauge.
So whatever its promise, digital instrumentation hasn’t yet produced any remarkable improvements in safety. Some of the reason may lie in the failure of training programs to keep up with the technology. Paradoxically, new students learning to fly in glass cockpits suffer if they’re allowed to let their involvement with the avionics outrun their ability to fly the physical airplane … while experienced pilots making the transition to glass need to spend almost all their time learning to work with the panel, and preferably on the ground. These folks can probably learn to land but may need a while to master the details of diverting to an alternate airport and loading the required approach—while students training ab initio in glass may be able to pull up METARs from any airport in the hemisphere before they’ve learned to take off, land, or go around. It’s a brave new world, all right.