737 Max: 1960s Design, 1990s Computing Power and Paper Manuals

To be fair the article doesn't critique using 1990s (or earlier) hardware/software. It's merely a clickbait headline. The only reference is dismissed by the very next sentence as insignificant [Boeing could be replaced with any ~~expert~~ somewhat-knowledgable person on the subject]:

> The flight-control computers have roughly the processing power of 1990s home computers. A Boeing spokesman said the aircraft was designed with an appropriate level of technology to ensure safety.

This only adds to my lowered standards when approaching NYT articles over the years. They've long been abandoning being correct for being entertaining at a seemingly accelerating rate. Which is too bad considering plenty of people continue to hold them to a higher credibility over the typical clickbaiter sites. Considering they are a subscription service I'm not convinced this is justified.

The only thing a more modern processor buys you is more memory, a faster tick rate for your RTOS, potentially an MMU if it doesn't cause substantial timing skew, and high-precision floating point. In the case of an avionics system I would imagine the requirements for fixed-point arithmetic are very well-understood at this point, so floating point doesn't buy you a whole lot and potentially creates new and exciting problems.

So I would say that 90's-era computers are plenty powerful for what they're used for.

For example if a UI display showed:

"Warning: MCAS overriding pitch up. Disable?"

Is better than a LED #234 and LED 412 are on. Consult manual for instructions.

I would imagine that a lot of the fly by wire or control systems nowadays are either analog systems or very very simple digital systems (with extreme redundancy).

I think the way you criticize the article is about as faulty as the article itself.

This piece is about detailing how the 737 is an outdated design that got patched and patched till it more or less broke. The processing power is just /one/ aspect and is given as a point of reference, like the 1968 photo.

I am no expert in aviation, but I guess it's quite a safe guess that fly-by-wire systems need more processing power than mechanical/analog designs. And that a glass cockpit with a trouble database needs more than a blinking LED and a paper handbook.

Nowhere in the article does the author claim the 737 MAX crashed because it had too weak CPUs. It is, however, one indication of an obsolete design - newer jets like the A320 or the 787 have more beefy processors for a reason.

I know I worked in 1-dimensional transportation and their 486-based onboard-processors felt cramped, the software engineers had to hack around their limitations and they wished they could switch to something newer. I doubt 3-dimensional transportation is easier to implement.

There exist possibilities for far more safety with more compute.

For example, when a mechanical failure occurs (for example an engine explodes and partially falls off) the flight characteristics of the plane change dramatically. Current systems fall back to human control, and hope that human will be able to figure out how to control a now very "unique" machine.

Future systems could dynamically create new flight models based on collected data to be able to fly the machine in entirely new ways.

For an example of this, did you know it's possible to fully control a quadcopter even with 3 out of 4 rotors broken and no working fins/flaps? [1]. That sort of 'fly it how it has never been done before' is out of reach of human control, and might save lives.

[1]:. https://www.ethz.ch/en/news-and-events/eth-news/news/2013/12...

I don't think so. You build a plane that's expected to run for decades -- if you put a 2010 computer in there now, it'll be 20 years old in 2030.

That 1990's computer was put there in the 90's, does that really make it more reliable?

According to wikipedia, the AMD29000 processor was first released in 1988 and "In late 1995 AMD dropped development of the 29k..."

I'm not commenting on whether that's appropriate to be running your aircraft or not, but it does very much seem to be a 1990's CPU running approx 50mhz.

And paper manuals? Are they expecting to use iPads for in flight documents or something? Shall we compare failure states of paper vs tablets?

The classic approach is to have three sensors, so in case one fails you can know which one. Having two only indicates something is wrong but is not useful on the fly.

Are you saying the answer in the article didn't give enough information?

"Airbus addressed this potential problem on some of its planes by installing three or more such sensors. Former Max engineers, including one who worked on the sensors, said adding a third sensor to the Max was a nonstarter. Previous 737s, they said, had used two and managers wanted to limit changes.

The angle of attack sensor, bottom, on a Boeing 737 Max 8.CreditRuth Fremson/The New York Times “They wanted to A, save money and B, to minimize the certification and flight-test costs,” said Mike Renzelmann, an engineer who worked on the Max’s flight controls. “Any changes are going to require recertification.” Mr. Renzelmann was not involved in discussions about the sensors."

When you're nose down heading into the ground, taking your hand off the stick to adjust the trim can get panic-y. Especially since the MCAS may be working against your trim adjustments.

The design of the Max line seems to be just enough compete with the A320neo, and just enough to not "need" re-certification. But these two justs, together, may have cost several hundred people their lives.

Those other Boeing jets required costly new training. Boeing were trying to build an update to the 737 that did not require this.

Sometimes this is justified -certainly the MCAS incident is a example. Changes like the MCAS should never have been allowed to change a fundamental part of the airplane interface without considerably more processes being met. That said, sometimes opposition to enhancements is trade unionists being trade unionists.

Apologies for not finding an exact link, but here is the wikipedia article for the flight: https://www.wikiwand.com/en/US_Airways_Flight_1549

Even on aircraft without fully integrated checklists, many operators have switched to using iPads instead of paper manuals as well.

> A second electronic system found on other Boeing jets also alerts pilots to unusual or hazardous situations during flight and lays out recommended steps to resolve them.

The 737 is stuck in the past because given the choice, nobody wants to have to retrain their pilots on more modern systems.

Now, in the Air France flight, part of the problem was the system trying harder than it should have: it would have sufficed to alert the pilots to the lack of air speed indication and let them figure out what to do, but instead the computers cascade alarm after alarm.

You can see the problem: various pieces of the system were engineered separately, and there was no single system that could have suppressed the downstream alarms so that the pilots could focus on first on understanding the first alarm. Then again, that too might not have been a good design: perhaps a stall alarm should take precedence, say, over a frozen pitot alarm.

Within days of the Lion Air crash, MCAS was strongly suspected to be the problem. And yet Boeing management allowed 157 more people to die in the Ethiopian Airlines crash. And then the Boeing CEO still tried to tell everyone that it was a safe airplane.

The entire top management of Boeing, and the entire board of directors, should be perp-walked out the door.

Of course, that will never happen.

Because as we are literally discovering, not knowing the details of how and why the automatic stabilization software is driving your trim wheel can kill you. The biggest single part of the failure chain here was the fact that this should have been an invasive change requiring recertification as a different aircraft and retraining for pilots instead of pretending that it was just like a 737-600.