photo by Nathan Coats
IBM introduced the Personal Computer/AT – AT stood for “advanced technology” – in 1984, two years before I bought my first computer. The product locked down Big Blue’s position as the dominant personal computer maker, which in turn made Microsoft a behemoth and Bill Gates the world’s richest man.
The processor at the heart of the PC/AT was the Intel 80286 chip.
Boeing introduced its 737 MAX series of jets in 2017, one-third of a century after the PC/AT. The MAX immediately became Boeing’s biggest-selling aircraft and seemingly locked down Boeing’s position as the world’s largest passenger aircraft maker. That is, until two tragic crashes five months apart led to the MAX’s ongoing worldwide grounding.
The processor at the heart of the 737 MAX jet is, startlingly, the Intel 80286 chip.
The 80286 is a microprocessor that was two generations obsolete in personal computers before my recently married daughter was born. And it may be a big part of the reason why Boeing’s difficulties getting the MAX back in the air seem to keep snowballing.
Information about the 80286 and its role in the MAX was reported in late June in a blog called Moon of Alabama, a quirky, anonymously written website that has been linked to a former pseudonymous writer for the Daily Kos. My skepticism was somewhat mollified when the report was picked up by Hacker News, a social news site run by the famous Silicon Valley startup incubator Y Combinator. They know their chips out there. If this report is accurate – Boeing has not disputed it as of this writing – then it may explain a lot.
In particular, it may explain why Boeing engineers designed the MAX flight control software known as MCAS (Maneuvering Characteristics Augmentation System) to rely on a single sensor to detect when the aircraft was at risk of an aerodynamic stall, even though the aircraft has two such sensors. MCAS was designed to force the nose of the plane downward, the correct maneuver to regain airspeed and stay aloft. But both the Ethiopian Airlines crash in March and the Lion Air crash the previous October, which together killed 346 people, are believed to have occurred due to a sensor malfunction. When the sensor that MCAS relied on falsely indicated an imminent stall, the system pointed the speeding planes back toward earth. In spite of the pilots’ efforts to keep their aircraft level or climbing, the malfunction had tragic results both times.
According to the blog reports, MCAS may have relied on a single sensor simply because the 80286 chips at the heart of the 737 flight control systems – on all models – are operating at or very close to their data processing capacity. Input from two angle-of-attack sensors might have overloaded the chips and brought the entire flight control system to a halt.
This may also explain why Boeing’s recent efforts to fix MCAS with a software update encountered a snag. Government flight testers, using simulators, found that the aircraft was unable to respond quickly enough to pilot commands in certain emergency situations. The problem was widely reported after Boeing disclosed it in a securities filing.
We still don’t know whether further refinements to the software will correct the problem or if the issue is the chip itself. Replacing the 80286 on the 500 MAX aircraft that have been built, as well as those that are on order, could be a major and time-consuming undertaking. Even beyond the hardware fix, this change would involve getting the new hardware and its software certified as airworthy by American and global regulators. All this could push back the MAX’s return to service by months. Some analysts told The Wall Street Journal that they think that it could even be years before deliveries of new MAX aircraft return to normal.
Against this backdrop, the news that I originally planned to write about today pales in importance. Last week Boeing lost its first major order for the 737 MAX to rival Airbus when a Saudi airline canceled its plans to buy 50 of the MAX models. The cancellation came despite the fact that the competing Airbus A320neo model is on back order, with a reported five-year waitlist. Undelivered MAX aircraft, meanwhile, are piling up at Boeing’s Seattle-area facility. Boeing has taken to storing brand-new but grounded jets in an employee parking lot.
All this points toward the idea that the original sin of the MAX design was that Boeing rushed it to market to compete with the A320neo. Primarily for marketing reasons, Boeing chose to make the aerodynamically altered new model appear to behave as much like earlier versions of the 737 as possible, to avoid forcing airlines to put pilots through lengthy and expensive simulator training.
The fact that an aircraft uses old technology like the 80286 microprocessor is not necessarily alarming by itself. The ancient Intel chip – which is now made by other companies – has limitations, but its performance parameters (including those limitations) are extremely well understood. That is crucial in applications where reliability is paramount. You don’t want passenger jet pilots to encounter the old Microsoft “blue screen of death” while seven miles high in the sky with 150 passengers or more on board.
Boeing made an excellent strategic decision some years back when it spurned a new super-jumbo to compete with the Airbus A380, which ultimately struggled in the marketplace. Boeing instead made a hugely successful medium-large wide-body, the 787 Dreamliner. But its attention to the Dreamliner seems to have left it blinded to the growing demand for longer-range single-aisle aircraft, until the announcement of the A320neo focused its attention to that unglamorous but crucial workaday market. This, in turn, led to the rushed timeline to develop the MAX and the marketing-driven compromises that, by all indications, have turned tragic.
Once again, we see the truth in the air safety maxim that serious accidents are almost always the culmination of a chain of events or mistakes that could have been avoided, at least with the benefit of hindsight. It is why accidents are so carefully reconstructed and studied, so the lessons bought at such a high price do not go needlessly to waste.
Unfortunately for Boeing, there may not be a quick fix for the MAX. Every one that has been tried so far has led to new discoveries that entail further fixes. Getting the MAX back in the sky is proving to be a very expensive undertaking. But at least Boeing is now paying the price in time and dollars, rather than lives.