I wrote before I read what you posted. I agree.

Fully autonomous is what arguments generally jump into too quickly. As mentioned before by others and myself I would be more concerned about the force reduction measures already in play. Fortunately nothing we have has been originally designed to take out the right seat like the shelved 797 was at least for now, not that a “patch” couldn’t be figured out however unlikely. It will be long haul first 4 to 3 pilots or perhaps directly to 2 with a ground monitored “Otto Pilot” box in the right​​​​​. Definitely a high mountain top to attain fully autonomous, but it’s the journey there which should be the main concern now. Our fleet runs bare minimums on bells and whistles vs the Fred Smith Faction and of course pax carriers always broadcasting/sporting the latest and greatest which always debuts first unless requirement driven. Upgrades haven’t been really at the forefront of ACMI. Replacement is one thing, upgrades are another. Those with less than a decade remaining probably have no worries and those with two may just have a totally new environment to contend with. Wi-Fi is a plus, latest tech with adaptable upgrades or transitioning software is a negative in many cases for our butts in seats. Just spitballing overall, but there will be significant impacts prior to full automation whenever that may ever happen especially here.

Agreed. My theory is that a lot of our discussion is really grounded in tort law. As long as it's expensive to kill or injure Americans there is a baseline level of safety that's driven by insurance companies and financial risk.

In regions where it's less expensive to kill people we see higher accident rates despite having similar equipment and procedures. The incentive to make a shortcut here or there isn't balanced by the potential costs of those same shortcuts. Same for costs due to lapses in oversight.

So liability and law factor here. Perhaps more than technology? I think this is why there are still people in locomotives.

You're assuming that automation is not safer than manned operations. That is perhaps not a correct assumption at least according to an analysis of human error vs. automation errors in aviation accidents. Large jets have been flying completely autonomously for decades. I have shared airspace and conducted tests with their operators and engineers. TCAS combined with control logic and current navigation technology along with a person on the ground to monitor systems equals a very mathematically safe operation. We have hardware limitations with current aircraft, but the technology already exists in robust form, we just need to implement it. I won't speculate on the timeline, but it won't be in the immediate future.

You're missing the largest piece of the puzzle. There is no current datalink technology that would be sufficient for the requirements of autonomous flights. The technology does not exist at the moment. As a matter of fact, it is likely at least a decade away.

I disagree. Google "Global Hawk". Yes, different mission and they don't fly out of LAX, JFK, etc, but they could. The datalink is robust, but nothing special. The technology is there, it just needs to be tweaked for freighter or passenger ops. Certainly single pilot ops is very doable.

Which large jets has been flying autonomously for decades?

Ok, my definition of "large" was relative to small UAVs. It is the Global Hawk, a pretty big jet. But tell me, what about this technology prohibits it from applying to freighters like the 767 etc... The Air Force has flown everything from fighters to Boeing aircraft without any humans aboard, just not in the NAS. The technology is there. It's surprising to me that people don't know this.
Google Boeing 720 remote piloted Edwards AFB

Global Hawk is pretty big. Like you, I shared airspace with UAS for years. There are lots of us here. I have personally watched them fail and bust across boundaries, disrupt stacks of assets working a particular task, etc. I have personally watched manned assets do the same thing, though. So I don't assume UAS or automation is inherently worse; it's inherently different. Crews have strengths and weakness (fatigue and cabin environmental factors, for example) and automation has strengths and weaknesses (Iranian capture of an RQ-170 for example).

I was really trying to speak about liability, economics and politics more than technology, though. That's why I point out that trains still have conductors even though automated trains may be more reliable. If we aren't willing to chance automated trains the limiting factor is law and politics rather than technology. I didn't make my point very well, though.

Ten years ago I was telling everyone we'd see single pilot ops in ten years... So I'm saying it again, just give it ten more years, eventually I'll be right