Automation Is Making Us Stupid

I keep asking myself how it is possible that airline pilots are landing at the wrong airports. It seems inconceivable to me. I became a pilot when training was dominated by analog systems. Apparently, pilots are now being trained to rely on digital systems to keep them, and their passengers safe.

Many readers have heard that the Boeing 737 Max 8 has being grounded by almost every country because of questions about its ability to remain in the air after two recent crashes. There was some controversy because the United States was the last nation to suspend the flight operations for the Max 8.

What caused the crash of the two recent 737 Max 8 aircraft is yet unknown. Like all pilots I have my opinion as to what could be the issue the led to the crashes. But it is important to note that I do not have access to the crash data, or black boxes and the little information I know about the two incidents comes from the news media.

However, I have experience in the 737, specifically the so-called “classic” or the second-generation of the airplane. The three productions series (300/400/500) of the classic entered service in 1984. The Max series did not enter service until 2017.

There is a significant difference between the Classic and the Max.

Pilots are “type-rated” to specific aircraft. In other words, among the type-ratings I hold includes the 300 series. I flew the 737-33A. The reason ratings are required is because all aircraft have different flight requirements. These include start-up & flight procedures, emergency recovery procedures and special procedures for that aircraft. Type ratings are attached to the pilot’s license.

When I was tested for the 737-type rating the instructor expected me to know, by memory, what items were powered by the “A” hydraulic system. This is important to note in order to deal with in flight emergencies. This is in addition to things you can look up on tables in the aircraft’s manual before flying, like the relationship between outside temperature to the configured weight of the aircraft.

Flying the aircraft during the type-rating training focused on emergency recovery scenarios such as engine out or other mechanical failures.

My 737 experience can best described as the analog version of the aircraft compared to the Max versions. Although the 300-series had advanced avionics and autopilot, in my opinion, there was much more manual control of the aircraft, especially during take-offs and landings. The thing I remember the most about my 737 experience was that the aircraft tended to be too sensitive to pitch up or down compared to other aircraft I had flown.

This is where I feel that the issue lies in the two recent accidents.

As new aircraft are introduced new disaster prevention technology is added to the aircraft. New technology leads to computers making decisions instead of the pilots. Push the nose too far down you crash. Pull the nose to high up you stall and likely crash if your altitude does not permit time to recover from the stall.

To prevent takeoff stalls, technology has been introduced overtime to remind pilots to properly configure their aircraft for takeoff. For example, extending slats and flaps to add more surface to the wings for slower speed operations. I have seen crash reports where pilots forgot to extend the flaps during their takeoff checklist leading to the crash.

As basic as that sounds, especially when procedures are followed where a pilot checks off the flap item on their takeoff checklist by visually checking the flap setting, the routines of multiple flights leads to pilots just reciting from memory instead of physically checking the items on the lists.

Aircraft manufacturers added klaxons and master warning lights to alert the pilots to a misconfigured aircraft, but distracted pilots ignore the basics and focus on complex failures leading to wasted time looking for a failure that does not exist. There is not much time to recover during a takeoff and thus flaps not properly extended will likely end in a crash.

To alleviate this, manufacturers have started to add computers to assert control over the airplane to recover from emergencies. One of these is an automatic attitude/pitch control that reacts to a stall and pushes the nose down to recover the aircraft from the stall.

The problem, though, is that the attitude controller is dependent on another piece of technology, usually the attitude vane, that tells the computer the pitch the aircraft is in. If the vane is malfunctioning it might indicate the wrong pitch to the computer and the pilot. The computer only knows what the technology tells it. The pilot, however, has more information to work with. The pilot’s instruments might be telling them that their aircraft is in a 30-degree pitch up and on the verge of a stall, but the pilot’s senses tells them that the aircraft is not in a stall and that the pitch appears normal.

In an analog cockpit the pilot makes the decision to stay the course and continue to climb regardless of the audible warnings and lights. In a digital cockpit, the computer takes over to correct the assumed stall and pitches the nose down forcing the pilot into a tug-of-war with the computer pushing the nose down and the pilot trying to level off the aircraft.

Very early indications are that the computer is making decisions for the pilots. Maybe the pilots weren’t trained properly to override the computer, or maybe the computer took too much control. It is also possible that there was just not enough time to diagnose the problem and reach a solution. We won’t know this until the final reports are out. It could be years.

But when you tie in the fact that pilots are landing at the wrong airports because of their reliance on technology makes it very plausible that we’ve reached the point where the question that must be asked is whether it is the pilots in command, or the computers?

Remember that we are on the verge of mass deploying autonomous cars on the streets where the computers decide how to drive. Is this really a good thing?