It's been more than two years since the shot-in-a-million collision over Brazil of a GOL 737 and an Embraer Legacy 600 that killed all 154 people aboard the Boeing. The Legacy was damaged, but the crew landed it with no injuries to any of its occupants. Immediately after the accident, Brazil outraged the aviation world by criminalizing the accident investigation and detaining the two Legacy pilots for more than two months before finally releasing them.
If anyone thought cooler heads would prevail as the accident investigation matured, they would have been disappointed by last week's decision by CENIPA, Brazil's accident investigation bureau, to cite the Legacy crew's negligence as a probable cause when clearly the report fact pattern shows that ATC ineptitude and lax procedures were primarily at fault. That's not to say the Legacy crew had no role, but the overall responsibility for separation of IFR aircraft is ATC's responsibility. Period. And Brazilian ATC failed to do that.
The CENIPA report made an issue of the fact--or at least claimed--that the Legacy crew accidentally switched off its transponder and suggests this was possibly due to minimal experience with the airplane and its avionics, plus poor CRM that distracted them from noticing. Also, the two Legacy pilots had never flown together and had little experience in the region.
It's a fair point to make. Had they been more intimate with the Legacy's glass, would they have noticed the transponder in standby mode? Either way, had the transponder been on, it's likely that the approaching GOL 737 would have gotten a resolution advisory in time to initiate evasive maneuvers. But again, that's a side issue. As far as ATC's responsibility to separate traffic is concerned, TCAS might as well not exist.
Another issue that can't be so easily dismissed is lost comm procedures, where there's a cultural divide between the U.S. and the rest of the world. The Legacy had lost contact with its controlling center, something that's not unusual in the Amazon region, where comm coverage is intermittent. But the loss of comm went on for nearly an hour. In this case, a workable comm link probably was available, but due to a cascade of ATC errors, it was never established. Similarly, other aircraft were on the frequencies the Legacy tried, but no relay was attempted. Although the Legacy wasn't in continuous radar contact, the lost transponder was detected but not acted upon by ATC.
In the U.S., the lost comm expectation is to remain at the last ATC-assigned altitude, terrain clearance issues notwithstanding. It's right there in FAR 91.185. Elsewhere in the ICAO world, the expectation is less well defined, but adhering to the flight plan has more primacy. In the Legacy's case, that called for a descent after a turn to a westerly heading that required an even rather than the odd cruising altitude that the Legacy had been on. This would be counter-intuitive to a U.S.-trained pilot. But the fact that many U.S. pilots aren't even aware of the ambiguity appears to have led Brazil's CENIPA to charge the Legacy crew with being unfamiliar with Brazilian regulations.
That's a stretch, in my view, but the point is still valid: Many U.S. pilots are so accustomed to the warm womb of full radar and radio coverage that they sometimes don't have the vaguest notion of what's expected when either or both are unavailable. Worse, they're not even aware that such cultural differences exist. (Exception: those seasoned international crews whose survival depends on knowing such fine points, not to mention variations on the theme due to local custom.)
In its appendix to the Brazilian report, the U.S. NTSB noted that the Legacy crew wasn't as aggressive as it should have been in resolving the ambiguous comm situation, this despite the fact that it tried to contact ATC 19 times. Still, the vastness of the blue sky all but assures a safe outcome following lost comm or radar. But as the engineers say, sometimes 10-to-the-9th comes early, as it surely did for the Legacy crew. The NSTB disagreed with the Brazilian claim that the Legacy crew's poor flight preparation was a contributing cause.
This accident reveals a seamy underside of the glass cockpit revolution and that's this: Pilots routinely launch behind glass technology that they haven't mastered. And there's not much choice to do anything else. However good the best glass training is, the systems are so complex and, in some cases, badly engineered, that it takes hours to master the routine and many more hours to adeptly handle the novel, never mind just flying the airplane and dealing with systems other than avionics. Too bad glass cockpits don't have the equivalent of Control-Z.
That the Legacy crew became distracted in fussing with their avionics is less surprising than it is expected. A few years ago, a friend of mine who's a sim instructor for a major airline, sent me a tape of a pair of line-qualified captains completely bollixing a simple approach procedure setup on an FMS. Between the two of them, they simply couldn't make the thing play right. My guess is that this is not a widespread lapse, but it's not unusual, either.
In this sense, pilots are no different than the population at large. Some are tech oriented and take to glass naturally, some adapt slowly and get by and a fraction shouldn't be allowed near it. That's no judgment call for or against the Legacy pilots, merely an observation about the state of training and technology in the industry. If passengers in the back think the guys up front are always all over those impressive displays and controls, they should understand that fake-it-till-you-make-it is still alive and well in aviation.
And now there's something new. This accident is probably the first GPS/RVSM-assisted collision, but mostly RVSM. According to the CINIPA report, thanks to RVSM, the two airplanes tragically achieved bullet-hits-bullet accuracy along the exact centerline of the airway, something that would have been unlikely with old-world VOR and baro altimeters. Had the Legacy been five feet lower, the two aircraft might have missed. Given the closure rates--about 1400 feet per second or the muzzle velocity of a musket--they might not have even noticed. As it was, the crew felt and heard the collision, but didn't know another airplane was involved.
Much has been written about the Legacy accident, but there's none better than this long article by the superb William Langewiesche. Here's the link. This piece contains links to cockpit audio recordings.