It’s no longer news that the avionics companies have done a great job of marketing glass panels, especially to manufacturers. As of this writing, Aviat still offers Huskies with analog instruments, and Diamond’s website suggests that you can buy a DA20 with a six-pack, but it’s getting increasingly difficult to find a new certificated airplane of any size with traditional instrumentation. Special orders are sometimes possible, but the published lists of standard features suggest that glass has pretty well taken over at Cessna, Piper, Cirrus, Diamond, and Hawker Beechcraft (not to mention Mooney before they suspended operations).
It’s still not clear whether this marks a real improvement in safety or just a triumph of salesmanship. The Air Safety Institute will offer some insights into that question later this year, but in the meantime there’s increasing evidence that glass contains traps of its own--not just for students and new private pilots at risk of getting swallowed by the screens, but for experienced instrument pilots who underestimate the time and training required to learn these systems well enough to fly them in instrument conditions.
Firm numbers are hard to come by because the details of system experience and transition training usually aren’t available for pilots in fatal accidents—in two-thirds of them, we don’t even know the pilot’s total make-and-model time—and more than three-quarters of all accidents in IMC are fatal. Some recent examples, though, show that a long history of flying behind conventional gauges is no guarantee that a pilot who’s new to glass won’t get in over his head, or that the consequences will be any less dire if he does.
A Cirrus SR22 crashed 4.5 minutes after taking off into a 200-foot overcast from Cuyahoga County, Ohio. During that brief, chaotic flight, the airplane’s attitude hit extremes of 50 degrees nose up and 60 degrees nose down. Bank angles reached 75 degrees. Airspeed varied between 50 and 172 knots and altitude from 320 to 2,320 feet agl; it apparently stalled at least twice and reversed heading three times. Data logged by the avionics suggested that the pilot, most of whose 400 hours of instrument time had been logged in a Beech Duke and who’d only taken the VFR Cirrus transition course, had apparently set the autopilot for “altitude preselect” rather than “altitude hold,” then failed to either reprogram the autopilot or disconnect it and hand-fly to a safe altitude before the airplane crashed. It also turned out that on his way into Cuyahoga County two hours earlier, he’d missed three consecutive coupled ILS approaches because he hadn’t configured the autopilot to capture the glideslope.
A Van’s RV-10 with a Dynon suite crashed near Seale, Ala., after its pilot was unable to intercept the final approach course for either the VOR 18 approach to Eufaula or the ILS 6 to Columbus, Ga. Altitude deviations during the last 14 minutes of the flight ranged from 400 feet above to 1,200 feet below the assigned altitude, and five times the controllers warned the pilot that he had wandered off his heading. A series of rapid turning climbs and descents in the last minute of the flight suggested spatial disorientation; most of the pilot’s 358 hours of instrument time had been logged in a Cessna 177B, and there is no firm evidence that he’d ever flown glass. With him was the airplane’s builder, who was not instrument-rated.
The pilot of a Cirrus SR22 that crashed into a retention pond near Indianapolis had bought the airplane just two days before. His 2,570 total hours included about 70 of combined actual and simulated instrument time in SR22s, almost all of it flown with conventional instruments. The airplane took off more than 300 pounds above maximum gross weight with its center of gravity aft of limits and the autopilot set for a constant 650 fpm climb, causing it to pitch up until it stalled. The pilot apparently failed to recognize the stall and continued to pull back on the yoke; the parachute was deployed too late to arrest the resulting spin.
None of these airplanes were rentals, but operators might fairly ask themselves what qualifications they’d like to see before letting a stranger take that new Entegra- or G1000-equipped beauty out in IMC (or even VMC at night). Clearly, total instrument time isn’t enough, and given the differences between competing systems, neither is time in some other glass cockpit. You could make a pretty good argument for requiring a check-out that includes a full instrument proficiency check in the specific airplane that’s being rented—it’s not a good sign if the pilot balks at demonstrating that level of proficiency, especially since a fresh IPC in the logbook never hurt anyone. Even that’s not a complete answer, however. The pilot of the Cirrus in Indianapolis completed an IPC in the very same airplane the day before the accident, and according to the instructor, flew very well.