Separate names with a comma.
Discussion in 'Hangar Talk' started by PeterNSteinmetz, May 28, 2021.
So he’s worried about a remotely controlled aircraft losing signal but yet he’s not worried about the autonomous software experiencing a “crash?” I guess it’ll be completely hardened against EMI effects?
This stuff is coming and my guess I’ll be obsolete in 15-20 years. Fortunately, I’ll be retired by then though. Not sure I’d want to be a kid these days with aspirations of becoming a pilot.
I suspect that if one uses 3 separate computers and voting the chances of a serious crash can be reduced far below the chance of communications failure. But I agree about human pilots likely being marginalized or eliminated eventually.
Isn’t there a science fiction story including pilotless vehicles on a bridge over the Strait of Gibraltar and the difficulties of getting passengers to accept that they were actually safer that way?
I think you'll be fine. And the next generation. From my perspective, the deciding factor how autonomous aviation will get on the commercial side, especially when carrying people, will not be decided by technology itself but who will be blamed when that technology fails and you get a body count.
If you look at the history of global aviation regulations and civil liability law its been following a steady path which in turn shaped the industry. One example of future could be how the 737 MAX issue finally plays out. But I think the benchmark will be set on new aviation technologies when the first commercial space flights take place and they have their first incident. Times are definitely changing.
I suspect it'll move towards management first (halfway there in the cockpit already). ATC will be similar, and a lot of the tools being put in place overseas are quite successful first steps. ATC has already embraced the ATM acronym to move towards management.
Not worried one bit, well beyond any time I need a paycheck. Even the ‘self-driving’ cars are off to a rough start.
I have 23 years left, and I am on the corporate side of things. I am not worried about losing my job to a computer.
This is the progression we'll see:
Big cargo to single pilot
Small cargo to autonomous
Pax airlines to single pilot
Big cargo to autonomous
Pax airlines to autonomous.
If there is one hiccup along the way, it will add years to the plan. One fatal pax crash and it could kill the entire thing, especially since we're on a really good run (12 years) since the last major US airliner crash. You will also have all the unions pushing back against this at the airlines, which will also add a few years to the process.
Plus, we still have freight trains with two people in the front of them. Till that changes we have nothing to worry about as pilots.
Every time I read one of these articles, I'm reminded of an excellent interview with the lady that runs (or ran) Duke's autonomous vehicles lab. She's a former F-18 pilot. She essentially said we tend to tackle the automation problem from the wrong end: We take care of the mundane stuff with automation, and still require operator intervention for emergencies and/or off-nominal behavior. Further, there's lots of studies that say human operators don't do a great job of monitoring systems for very long.
Her comments really resonated with me: Think of the skill, attention, and persistence you'd need to be on top of one of these things enough to be able to step in and "manage" it out of an emergency or off-nominal event. You'd be monitoring super-George all the damn time, and would have to be ahead of him in terms of all the ways he could f&%k things up for you. Far more workload than just flying the airplane.
A different, but very related topic was/is the idea that drones would revolutionize aerial operations (like ag/spraying, surveying, etc.) I've done a bit of work for a major aerospace company essentially "auditing" studies done to look at drones in these roles - and every single one of them was fraught with assumptions that favored drone operations (and deviated from the real world) and penalized manned operations After adjusting for these, the drone ops almost always "lost" the cost-effectiveness comparison with manned ops. I had a former student( a long-time CFI, a "student" only in the sense that I checked him out in an airplane he'd bought.) Like me, he taught on the side, and had a career in aerospace, with about a decade working for a successful military drone supplier. A startup was recruiting hard for their operation: using drones for forestry work, especially surveillance of forest health. We'd both come to the same conclusion: there was nothing this company was proposing that couldn't be done cheaper (probably a LOT cheaper) with manned operations. The cost of the pilot, especially a hungry young comm/inst/CFI/etc. is minimal compared to all the other pieces.
Between the Duke autonomous ops lady's comments, and the forestry drone startup, my gut tells me that real change is a ways off. That said, one of these startups may get lucky and hit a sweet spot - and then things could change more quickly.
Very interesting point about the cost-effectiveness. Perhaps for things that humans enjoy doing, you can get them to be cheaper than the required equipment to perform at the same level, even waiving the costs of the risks involved. This would seem to apply when risk to others is not a factor.