Imagine you engage your cruise control after leaving the gas station and entering the highway. Suddenly, the cruise control is behaving erratically, taking the car up to 90 miles an hour; what would be your response?
(Hint: Turn off the cruise control).
So number one, adequate training to allow you to turn off the cruise control before you encounter a bridge abutment.
Number two, do not let software control the cruise control. Would you let a phone app control your car?
Really appreciate the experienced and informed responses to the comment. Glad to see there is still some human intelligence out there.
I’m not saying that couldn’t happen, but, given the two-person cockpit (eschewing the traditional three up front), is leaving only one in the cockpit allowed these days?
Given the situation where you are operating a device that can kill two hundred people at once, it would seem obvious a human should be at least on standby at all times.
May be safe, but if it relies on mechanical inputs, there is always room for disaster. We need to stop relying only on automation to safeguard human lives. There has to be a backup.
It seems that human safety should be the last thing to be automated out of existence.
It’s worth remembering that ‘software flying the plane’ is a matter of degree; and one that all large planes of recent(I think the A320 generally gets name-checked as first to fly-by-wire and that’s mid 80s production accepted into service late 80s on) vintage participate in to varying degrees.
You aren’t going to manipulate the control surfaces or engine inputs of a plane that size manually; and electromechanical or hydraulic linkages are heavy and complex(and not necessarily designed for perfectly linear transfer of pilot intent from controls to control surfaces, you can do all sorts of damping and feedback and such in analog systems).
The difference is where on the scale from “lighter replacement for a hydraulic line” to “takes your input as a suggestion” the computerized systems fall.
Some are more or less literal, or at least don’t attempt anything notably less predictable than the analog systems did.
Some substantially decouple the pilot from the ability to waggle any specific control surface as he wishes; but do not interfere with your choice at a high level(‘climb’ might involve setting a bunch of different control surfaces and engine settings that you don’t get to control directly; but when you say ‘climb’ it sets them into the climb state for you).
Some use sensors and feedback to attempt to compensate for circumstances and modify their interpretation of your instructions accordingly. It’s this last category that seems to have been trouble here; with a system that both doesn’t seem to be reliable about it’s inferences and has an unfamiliar deactivation procedure compared to the earlier model, which isn’t a good combination.
Sorry, no. I am not a pilot, but my understanding is that these systems are necessary. The purpose of these anti-stall programs is to detect a combination of air speed and angle that is on the edge of causing stall, that pilots cannot sense. Without systems like MCAS, most of these planes would crash.
That’s exactly correct. Recall AF447 which stalled and crashed because the pilot kept trying to override the autopilot by pitching the plane up until it essentially fell out of the sky. That pilot thought they were in an overspeed condition which was caused by a blocked pitot tube but since it was dark out he had no awareness of the horizon. MCAS is a safety system designed to prevent this exact scenario.
MCAS is only activated with the autopilot OFF and flaps retracted anyway. Pilot always has primary control over all inputs and controls. Remember, these are extraordinary complex machines with millions of parts. It’s just not the same to compare it to a car’s cruise control feature. In fact, the sheer complexity of these planes requires automation because no human can possibly remember and monitor every setting in real time. We’ve made planes too difficult for humans to fly without assistance.
However, ask any pilot and their number one guiding principle during times of emergency - FLY THE PLANE! Remember, they are just as invested in a safe landing as the passengers are. There’s even a simple phrase that every pilot commits to muscle memory…#1 Aviate, #2 Navigate, #3 Communicate…in that order. But pilots are human and can be easily overwhelmed by too much confusing input.
We don’t know for sure what caused the 2 crashes so far but it almost always comes down to a series of mistakes, flaws, and decisions that combine together in a previously unknown way that lead to a catastrophic outcome.
Thanks, Mike. I can see you know something about flying planes. My argument is that pilots should be trained and ready at all times to take control of and fly the plane.
Arguing otherwise will support the claim that people are not capable of flying planes by themselves. The dynamics of controlling air flight has not changed since the Wright Brothers. Why are we making things so complicated that a trained pilot can’t take control of an aircraft and keep it from crashing?
When something is not behaving according to the laws of physics, it is probably a software problem.
Because flying is the easy - and boring part. It’s take off and landing that’s the hard part. Things don’t usually “just go wrong” in the air and ATC keeps a close watch on potential situations.
I’m pretty sure this isn’t true - but it is a fact that Trump was trying to get his personal pilot to lead the FAA. I don’t think he was ever confirmed.
You mistakenly assume that this is currently not the case. Commercial airline pilots are highly trained and have the ability to manually fly the plane at any time with total authority. Autopilots and other automation systems are designed to assist pilots - not replace them.
Unlike the Wright Brother’s flyer though, modern aircraft are extremely complex machines. Even if the physics of flying hasn’t changed the mechanics and methods certainly have. Systems - whether mechanical or electrical - are still human designed, human installed and human controlled. The point of failure here is the human - not the system.
We trust machines with our lives every day even if we don’t realize it. However, If you’re arguing that we strip out all the modernizations and software controls out of airplanes to make them safer I would say you are seriously misguided.
BTW- I’m not a pilot either but just an avid aviation enthusiast.
There are fighter jets that literally cannot be flown by a human pilot without software compensating constantly for the inherent aerodynamic instability of the airframe. As in, would spin out of control, crash and burn while still on the runway.
As long as you’re cool with planes that are no faster than a WWII-era fighter and no bigger than DC9, then that works. But for modern aircraft, you’re way, way off.
Okay, so why are all of these cases saying the pilots were fighting these systems designed to “assist” them in flying the plane?
Complex or not, pilots are the final defense in protecting the safety of the passengers. If there is anything that gets in the way of that, it should be eliminated.
Are you arguing that pilots are incapable of flying these planes?
Root cause: depending on software to insure human safety.
I’d guess combination of a brand new aircraft design and lack of extensive training on it. Apparently it “drives” much differently than 737s of old which could explain the fighting with the systems.
No. I am saying that pilots should know how to disengage these systems when necessary and be prepared to fly the plane manually. That is what we hire them to do.
