rickair7777

So robot planes are just going to pop up on departure frequency and check in with ATC? And that will be the first ATC ever heard of it? Get real... the federal government couldn't develop regs, certify planes, and modify ATC to handle it without a decades-long manhattan project. And somebody will have to convince the fed to spend billions and billions to eliminate a few good union jobs to make some execs richer.

And there-in lies the real problem... nobody can invest the massive R&D money needed to pull this off without government regs to guide their development a predictable timeline for certification. Chicken or Egg...

It will presumably happen eventually, but it will progress slowly over time, decades or perhaps a couple centuries.

Automated elevators have existed for 100 years, but the public did not accept them for another 50 years.

Trains are still not automated... could they be? Sure, like an elevator they operate in one dimension with some consideration for a second dimension (slowing down for curves). Why are they not? It's cheaper to pay someone than buy all of the automatic systems (with backups to the backups to the backups) necessary to replace the human. Even cargo trains still have drivers.

Also liability is a huge factor. Juries and the legal system understand unintentional human error, they'll mete out appropriate consequences. But they won't seek to bankrupt the vehicle manufacturer because some drunk driver ran a red light.

But an automated vehicle focuses 100% of liability on the manufacturer (with some possibly shared by a maintainer)... there will be no sympathy for a manufacturer which unleashes a menace on the unsuspecting public. Don't believe me? Look no further than MCAS... that's going to cost many billions, even tens of billions by the time it's said and done.

Now imagine an aircraft operated by 10,000 autonomous MCAS style systems...

And speaking of MCAS, autonomy is now that much harder at least in the US because anyone who's serious about it envisioned industry designing their regs and mostly self-certifying because autonomy and AI are pretty far beyond the FAA's capabilities. They basically would rely on "regulatory capture" to get their stuff flying. Well MCAS sure turned the tables on that.

I've said before that I had a suspicion that industry money-men might rush it, fail miserably and spectacularly, and set the whole thing back for even more decades.

I wouldn't worry too much as long as trains still have engineers.