The Lure Of The Cheap Autopilot
Anyone who transcends beyond casual skepticism understands that to be truly blackhearted is to know that cynicism is the smoke that curls up from burned dreams. Personally, I’ve embraced it, with a list of failings and character flaws so numerous that I need an Excel spreadsheet just to list them.
But one thing I am not is a Kool-Aid drinker. Oddly, my Catholic school upbringing saw to that; I perfected the eye roll by fifth grade. So five years ago, when the FAA was describing the coming wave of streamlined certification as “twice the safety at half the cost,” my BS detector went off-scale high. Lately, it’s come off the peg, but just barely.
Are we there yet? With regard to avionics, at least, we can measure the cost part. Let’s focus on autopilots, specifically the TruTrak Vizion system at an eye-opening $5000 that Andrew Barker talks about in this video we shot at AEA in Last Vegas this week. And that’s no stripped-down wing leveler either, but a full-featured autoflight system with envelope protection.
A decade ago, the least expensive autopilot installations were in the $15,000 range, if not a lot more. Even today, as TruTrak, Garmin and Trio offer capable autopilots for five grand, the Genesys System 30, the old rate-based S-TEC bargain AP, lists for $13,000. So we’ve hit the half-the-cost claim and then some.
Now the safety part. Generously, this is a rubbery concept. What’s “twice the safety?” My working definition would be a halving of the general aviation fatal accident rate. Despite the rosy motto, the FAA’s stated goal was to reduce the accident rate by 10 percent over a 10-year period between 2009 and 2018.
Confronting the data fairly, total accidents and likely the fatal rate have declined sharply during this period. In 2009, the fatal rate was 1.33/100,000 hours compared to 0.98/100,000 in 2016, according to NTSB data. The FAA claims a bigger drop, but I’ll stick with the NTSB data for the moment and to be fair, it exceeds the FAA goal anyway.
What happened? A combination of factors related to improvements in training, an aggressive attempt to teach risk mitigation, better maintenance management and perhaps some impact from more sophisticated avionics, mainly glass panels and autopilots, although a 2010 NTSB study on this very topic found no safety impact from the glass evolution. Eight years hence, perhaps the effect is more measurable. And I wouldn’t discount the impact Cirrus has had on the whole in reducing its own accident rate dramatically.
And that gets us back to affordable autopilots with envelope protection. Despite all the training and risk awareness, the big killers in GA remain loss of control and stalls. Because we can’t interview dead people, the reasons for LOC accidents are poorly understood, so the operative theory is that smart autopilots can intervene frequently enough to nudge the fatal accident needle downward. It did not escape me that Garmin’s new GFC 600H autopilot for helicopters has a hover assist mode, perhaps to help the ham-fisted helo driver avoid trimming trees with the tail rotor.
Certainly, in onesies and twosies, that ought to move the accident rate downward, right? It probably can’t hurt, but I’m skeptical that this technology will have measureable impact at this point.
As long as he’s in the loop, homo the sap is ever creative in digging smoking holes by defeating the very systems he invented to keep him from doing so. There’s no line of code in an autopilot that’s the equivalent of “Hey, watch this.”
What we’re missing here in the forest-for-the-trees wonder over impressive autopilots is that we’re well past the dawn of a sea change that’s marching smartly toward autonomous flight. You can’t help but notice all the coverage we’ve been doing on drones and automated flying machines that will whisk passengers from A to B at the press of an app option. Envelope protection is but a mere momentary stop on that road and, being a little harsh here perhaps, it’s already obsolete. We just haven’t realized it yet.
I’ve been doubtful of the timing and remain so, but the outlines are unmistakable. Line 266 of my spreadsheet catalog of character flaws is the inability to predict the future. But I’ll hazard a prediction anyway. The coming GA bifurcation will be between pilots for whom coping with the prospect of a fiery death is the appeal of flying and those who just want to fly to South Bend for Christmas; those who lust for the feel of the stick and those who admire turn anticipation of a perfect radius drawn on a vivid TFT.
In the not-too-distant future, someone will be blogging about how control laws should be written to balance the pilot’s genetic urge to intervene against the wisdom of preventing him from doing so.
Welcome to the revolution.