Why do you need to press so many buttons to start a plane? Can’t there be just one button to start everything in sequence automatically?
In: Technology
You want as much control over such complicated systems as possible, if only to be able to work through any in-flight emergencies if need arises.
And, modern aircraft are quite automated anyway, just compare modern computerized cockpit to one from 1960s. Some aircraft had a separate crew member to just control/oversee the engines and other systems due to super high workload.
It is still hard to fly a passenger liner compared to driving a car, bexause you have so many systems there (and often, most important ones are doubled, or quadrupled for safety, too)
When you start a car, the twist of a key actually automatically activates many separate systems. First it powers the electrical system to many critical components. Fuel pump. Engine control unit. Ignition system. Then if you continue twisting the key it engages the starter motor, and starts the engine. You release the key and all of the systems keep the engine running.
In an aircraft all of these critical flight systems are isolated on different switches, and often have multiple backups (fuel pumps, ignition systems, hydraulics, air pressure, different sources of power for the electrical system, totally separate electrical systems). There is a redundancy and level of safety here that lets a pilot troubleshoot and select or de-power systems individually in the event of a failure while in flight or on the ground. Hence the complexity of starting. It requires a certain order of operation to bring everything online to start a plane.
EDIT: Some seriously good discussion here about aircraft. A couple of points to make. Are we talking about starting the engines only, or taking an aircraft from cold and dark to ready to taxi? I agree that FADEC systems allow an automated start. But even then you still have to at least turn on the battery master. And even if the aircraft could be fully automated, that wouldn’t preclude the need to check and verify that everything is working and set appropriately. Also, automated systems are their own system that needs to be monitored with the possibility of failure (MCAS). I’m not making an argument one way or another, thanks for the discussion.
EDIT 2- One final comment guys. Many people have said this, and I will agree. A fully automated, one touch button start aircraft, complete with systems integrity monitoring is technically possible. And in small drones you can actually find this technology today. The question is, is this acceptable, is this safe, and how can we determine the risk going forward. Commercial air travel is by far the safest method of travel in the world today. It’s a really really high bar to even meet. Getting technology to match that level could be coming in the future, but it’s going to be a long and slow implementation.
In a car you’re responsible for 8~ lives and there are not as many systems in play as a B737-800 for say (a car engine dies, you slow and stop; a plane engine dies, you now have uneven thrust on a vessel with hundreds on board and “stopping” means hitting the ground, however well you’re able to do it), which can carry hundreds of souls. if you look through youtube for preflight checklists youll see how necessary each step is. one big thing is safety, make sure all your warning lights work (there’s a test switch for all warning lights). the other is airline cost, you dont want to start your engines until you need them (fuel cost AKA cost index), and before that you need to make sure everything is safe to function in the event of a failure. some planes like the 747 can’t even move until al 4 engines are running because if only 2 are running they would produce an unsafe amount of thrust to move around the airport to the workers on the ramp. so mostly it comes down to safety.
and then there’s redundancy. You can start or restart an engine with the APU, or with the bleed from an already functioning engine, or from electricity from either the apu, a generator, the battery, or ground power. with so many variables you have to check to make sure each is available to you in the event one somehow isn’t.
sticking with the theme of safety, you have to make sure your instrument readings are correct, setting the QNH (make sure your altimeter knows where the ground is), check that your settings are ready for takeoff (go full throttle for a second and make sure the warning sound beeps when you know you aren’t set for takeoff) etc.
a lot of this is programmed into the Flight management system (FMS/FMC) and in fact pretty much all of the flying is done by the computer once you reach the “transition altitude”. but the onus is on you to make sure the computer is doing everything right, and you need to know how to solve a problem if one arises.
Tldr: It’s more or less a checklist to make sure all systems are working properly.
Some of it may include adjusting the instruments on the dash so the readings are correct based on weather conditions and compass readings from the ground. Radio check on all the radios. Listening to the weather forecast and writing it down. Depending on the aircraft, you may need to prime the engine, start it up, check and make sure it acts as expected when you adjust things. Make sure back up systems are working well. Make sure all the control surfaces, flaps, evelons, rudder is working. Check the exterior lights work. Check to make sure all the sensors, radars, and other equipment is working. Hydraulics and back up hydraulics. Cabbin compression, emergency equipment. Navigation equipment.
That’s about all I can think of, but it’s quite a lot.
Almost every thing on a commercial plane is going to have main systems and back up systems. You want to check it all out before flying.
Even after the plane starts moving you have to check a few other things, like the brakes and throttle.
[Newer high-end airplanes](https://www.youtube.com/watch?v=qlR0mszVIng) do have automated start sequences. Technically it isn’t too difficult, it’s just that it requires a computerized sequencer to operate all the switches and valves based on inputs from other instruments. When you have critical equipment run by a computer you have to go through a painful, long, expensive process to qualify the software and hardware.
I work in nuclear. Software QA is similar — usually easier to deal with ancient instrumentation than find something with acceptable software QA.
There could be, but it would probably make the aircraft much less safe. It would be pretty easy to program a computer to go through the normal cold start procedure, even accounting for conditions that would require a the procedure to be modified, perform pretty much every diagnostic check that a human pilot would. It would be 100% reliable, under 2 conditions: every sensor is giving the computer accurate data, and the thing was set up correctly (both by the programmers that made it, and the pilots that pressed the button).
However, in the real world, you can never rely on both of those things being true. Having actual humans go through the means that even if technically all the numbers are right, they can investigate if something doesn’t “feel right”. This means they can catch edge cases that the programmers of an automated system might have missed, and also might notice if a sensor is giving bad data. Also, in the case of an emergency, it’s vital that the pilots are aware of how every system on the plane works and interacts with every other system. By manually switching power from the battery, to the APU, to the engine generators, they’re essentially reviewing how the plane’s electrical system works, which could be the difference between a safe landing and a crash if there is an electrical fault in flight.
I remember listening to an NPR story about this a while back. The pre-flight checklist (and the general idea of a checklist as we know it today) was invented by Boeing after two pilots died in a 1935 crash of a prototype plane in Dayton, Ohio. The system failure occurred because the pilots had forgotten to disengage the gust locks prior to takeoff, which are only supposed to be active while on the tarmac.
Basically, as others have said: even if a pilot thinks they know the switch is in the right position, the stakes are just too high to assume they are correct. Human error (or in the case of your original prompt: program error) is a thing, and hurtling through the air at 500mph isn’t something that should be taken lightly. Hence the double checking of every function on board to ensure safety.
It’s not just planes, most computer systems, factory plants and vehicles which are not made for the end user are a lot more complicated to start, maintain running (flawlessly) and shut down then a car.
As an example I can talk a bit about my work in a candy factory. The plant to produce the candy slurry has an estimated learning time of 6 months, and then you’re barely qualified to start it and shut it down, if nothing unusual happens. The process from the point of powering the controls up (pressing the power on button, so to speak) until everything is set for regular use takes about an hour.
There’s a similar gap in user accessibility between home owned PCs and servers used for companies and specialized tasks.
Part of it is the checklist aspect that others have mentioned. Also, it’s kind of like why high end cameras have tons of controls while your smartphone camera has very few and does most of the work for you. Cars, and smartphone cameras, are designed to be easy to use in most circumstances. They are highly automated because that makes them easy to use. The tradeoff is reduced control, which means that when you encounter one of the circumstances that the automated systems can’t handle well, you don’t have easy access to controls for the individual subsystems, AND you probably don’t understand them well enough to know what to change to make it work.
In a smartphone or car, that’s generally ok, because you can put the smartphone down, or pull your car off the road. On an airplane, it’s not ok. The pilot needs to be able to directly control many different systems at a moment’s notice in case something goes really wrong, and they have to be accustomed to directly controlling those systems, so that if they ever need to do it for real, it will come naturally to them. They might not have time to fidget.
So that’s the other part. Pilots need to be masters at controlling many different subsystems because a situation COULD occur where they need to assume direct control of one of those systems.
Case in point: the two recent disasters involving 737 MAX aircraft. There was an automatic system to handle keeping the aircraft’s angle of attack (how much its pointed up/down) at an optimal level, and the pilots weren’t trained in how to take direct control of that system, and didn’t have an easy way to bypass the automatic system (which malfunctioned) and assume direct control. Result.. two crashes, hundreds dead.
Airbus pilot here. We have 2 switches to start the airplane, the first is to tell the computer we want to start the engine and the second initiates the start sequence. I flew Boeing before and they have a more complicated sequence, but without going into too much detail the Airbus way is just better.
Feel free to AMA I’m off today!
Latest Answers