Airplanes long before cars, as they have several advantages. Unlike the road, the environment in which they move is controlled. Computers already fly most airliners most of the time. And fully autonomous flight is already commercially exploited. But again, not before the vast majority of airline customers would be confident with no (or just one) human up front, which won't be until the generations that were taught to be critical towards machines and computers expire.

Fully qualified experienced pilots can decide to crash an airliner as well, intentionally or not, and they did so much more than a computer ever did. In the end, it boils down to who you trust more to make split-second life-or-death decisions for you, and that is a generational divide nowadays.

You're going to be pretty hard pressed to find a generation that's going to be onboard with allowing a human engineered software system make split second life/death decisions on an airliner. We're already finding that using such systems for supplementary functions can be catastrophic.

And yet we're using and relying on them on a daily basis both in the air and on the ground. When they betray us (and they do, no denying here), we patch them, and they will never make the same mistake again. OTOH, we cannot patch human brains, and mistakes are repeated over and over. As a result, the visibility of pilot error in air crash statistics is growing.

When pilot error becomes dominant in crash statistics, who are we betraying when we refuse to find a solution? Should we go in the history books as the generation that reversed aviation safety statistics because they couldn't look further? We shouldn't, and a lot of smart engineers are making sure we won't.

You're contradicting yourself. You say we should not be relying on the human brain yet it's a human who designs the systems that are there to automate functions.

Pilot error has been a factor in flying ever since flying began. In recent decades, accidents related to pilot error have decreased significantly. The same cannot be said for accidents related to aircraft automation. Automation is a great tool to supplement flying an aircraft, however it is not even close to being reliable or adaptable enough to be applied as the primary means of operating an aircraft. In fact, it's not even reliable enough to drive a car yet (cough... Tesla... cough...).

Now we have an aircraft that has been grounded for months because a hasty introduction into service in order to keep pace with a competitor has highlighted massive deficiencies in the design and engineering of said aircraft as well as extremely lax oversight by regulators and you think we're ready to start letting these people engineer things that fly on their own with no oversight? Absolutely not.

A human brain comfortably sitting before a computer screen in a large, quiet, well-lit, air-conditioned room works vastly different vs a human brain that may be sleep-deprived, may have to deal with stress at home or on the job, and is suddenly confronted with an emergency situation where tons of bells and alarms are going off, that requires knowledge to be recalled, and that requires interventions that may go against "gut feeling" but are necessary. All this from the AF447 crash which I took as an example.

Respectfully disagree with what seems to be implied. The total amount of accidents has decreased. The percentage of accidents caused by pilot error has increased. And not slightly.

AERO - MEDA Investigation Process
https://en.wikipedia.org/wiki/Catego...by_pilot_error

It is. That's what drones do. I agree, however, it is not the case for a flight carrying human lives. But that's only because there's simply no demand yet. When demand comes - and it will - technology will follow.

In the end, these were all human errors, with a company (a group of people in charge) becoming too comfortable. I wonder how many actual Boeing engineers were confident with the path the company took.

Might be worth pointing out in autopilot the problematic MCAS function is/was disabled. It was for the human pilot situation MCAS was included (with it's presently known issues).

And autolanding systems have been used for years in difficult weather situations.

The first part I 110% agree with, MCAS was an effort to game the system and get a fundementally over-stretched design to fit within (a) airworthness requirements, and (b) a single type for certifications.

I 120% disagree with the second. Boeing looked a new aircraft at least three times. Each time rather than thinking of the future they played to the analysts and the market which wanted to limit investment and try to push up margins. Then Airbus got the jump (again) on them, and this was Boeings half-... response. It will end up costing far more than a new clean sheet design would have by the time this is all ended, and Boeing will be hobbled going forward.

And new high tech airplanes are quite popular. I keep flying the A220 on DL, it blows the doors off the c1967 7-narrow-7 with its 24" wide bathrooms, narrow seats, and jury rigged systems.

Delta rightly sees the A220 as a competitive advantage, as do airlines that have nice shiny new A321/A320neos. The problem lies with Boeing's management, and the US's awful system of judging companies by each quarters results, forcing them to manage for the short term hit to the bottom line, not the long term.

Boeing and United (and now AA) are sort of two short term peas in the same pod.

Boeing has stated that they are going to do a clean sheet design of a single-aisle aircraft. The MAX is merely a stop-gap solution to combat the A321/A320neo until the new AC is ready for market which takes a long time as we know.

Like Airbus got the jump on them with the white elephant A380? The A350 which came to market after the 787? I will admit the A220 was a smart move on their part but they didn't even develop it, Bombardier did.

Delta also flies MadDogs still, what's that saying about glass houses? I wasn't aware that the AC manufacturer was the one responsible for selecting whether the airline wants to squeeze in a few extra rows or not.

Outside of FFers and Av geeks most people don't really care about what plane they fly on. The MAX being a possible exception due to this fiasco.

A very long read but contains some experiences and perspectives that I don’t think have been explored here.

https://newrepublic.com/article/1549...ial-revolution

Somehow, I don't consider the New Republic to be the bastion of investigative aviation journalism.

Based on the title and first paragraph alone, this is obviously just another political hit piece.

You took the time to respond, maybe read it.

It takes FAR longer to read than the four minutes between me posting it and you replying. Why comment at all?

Because I was intrigued by the idea that it might actually have new insight, and then I quickly realized that it was a political hit piece, exactly as I said it was. There was no need to read further, because even if there were valid points to be found, it was going to be extremely difficult to extract them from the taint of the surrounding material.

Aviation safety shouldn't be political.

Agree. Way too much political (anti-capitalism) spin in that piece. Regardless of which side one is on, credibility of the article vanished not far in to it IMV. Please give me facts, not opinion (esp non aviation opinion). 👎

I read it and there wasn't really anything that hasn't been covered in the first 150 pages of this thread. It also got some significant technical and organizational details relevant to the crashes wrong. All the evidence suggests that the software was written correctly (as in met the specification that was supplied) by the low-cost, outsourced programmers, and that the control laws that they were given (or that were applied in an update) were the issue. It was most likely more of a system engineering and configuration management issue, not a cut-rate programmer issue (at least as far as the available information so far indicates).