48

Looking at this picture of an A380 cockpit I see that there are still a lot of traditional switches (meaning manual click on and off or knobs). Being technology so advanced in these days, why is it that traditional switches and knobs are used instead of having a couple of touch screens on which the various controls are placed?

Wouldn't this make it more economic as there would be savings on material and wiring? They could also be less susceptible to failures as traditional switches and knobs still rely on mechanical actions.

Another advantage I could think of is that every pilot could customise positioning of controls and switches based on his/her way of piloting. This could then be reset to a default state every time a new pilot takes over.

enter image description here

Manu H
  • 16,468
  • 25
  • 101
  • 210
Fabrizio Mazzoni
  • 9,882
  • 15
  • 64
  • 113
  • 32
    Just look at the big mess all home equipment is now. You never know how to use it without opening the manual, or you keep clicking 23 times like a mentally ill just to change from summer time to winter time, and you are lucky if you manage to stop clicking before being too far :( – mins Nov 06 '15 at 13:04
  • 14
    As for any electrical systems, you should think of what happen in case of (partial or total) electrical failure. Also note that for home devices, keyboard are easier and faster to use than touch screen for many usages (e.g. post on this website) – Manu H Nov 06 '15 at 13:27
  • 3
    Closely related – Pondlife Nov 06 '15 at 13:48
  • 1
    @mins There are many manuals, and you'd never know how to use a modern commercial jetliner without years of training and studying of many types of reading material. – Todd Wilcox Nov 06 '15 at 14:17
  • If electricity where to die out on the aircraft for whatever reason the screens would shut off, so you would still want those switches just in case of that kind of situation. That's just my thoughts on it though. – Ethan Nov 06 '15 at 14:21
  • 14
    @ToddWilcox: But in a B737, at least when you want to change the bank angle limit you don't have to remember that it buried in the configuration -> current flight -> FMS ->other -> angles -> bank -> max option. – mins Nov 06 '15 at 14:24
  • 11
    Already covered over on UX.se. – MSalters Nov 06 '15 at 16:09
  • 7
    "every pilot could customise positioning of controls" There are some fairly good reasons for the control layout to be fixed. Any pilot who is qualified on the aircraft should be able to sit down and know where everything is. No "reset to default" methods, no software glitches getting you stuck on the last pilot's weird settings, no accidentally moving an important function to somewhere unhelpful (or removing it entirely?) - particularly if you don't realise you've moved/removed the function until you need it in flight. – anaximander Nov 06 '15 at 16:53
  • 26
    You must look at what your hand when using a touch screen in order to get the correct spot. With a mechanical control, muscle memory will get your hand to the right spot, and the control shape (for major controls), will confirm that you're there, all without ever looking in the cockpit. For example, I can find the power button on the TV remote without looking at all, while using my phone as a remote for Kodi requires that I look at my phone - if I drag my hand looking for the pause button, I'll hit all sorts of other stuff. – FreeMan Nov 06 '15 at 17:44
  • 59
    "They could also be less susceptible to failures as traditional switches and knobs still rely on mechanical actions." As someone who designs devices using both touchscreens and physical buttons, all I can say about this is no. – reirab Nov 06 '15 at 17:49
  • 9
    I'm not a pilot, but I have worked in recording studios which could possibly be the only place where a person is surrounded by more knobs, switches, sliders and indicators than a cockpit. Smaller studios have largely switched to digital workstations where all of the controls are on a computer screen. There are a lot of advantages, but anyone who does a lot of serious work with them will eventually break down and buy an interface so that all of the virtual knobs and sliders are controlled by physical ones. Touch screens just don't create the tactile feedback the human brain works so well with. – TomMcW Nov 06 '15 at 19:30
  • 4
    @TomMcW: Makes sense to me too. We have 2 hands and 10 fingers, using a touch screen is a waste of human physical capabilities. Using a mouse is already a problem, most well trained computer users replace the mouse by keyboard shortcuts that are faster (e.g. Ctrl-C Ctrl-V). Forcing such users to use only the mouse or the finger and the Edit menu creates an important frustration. This is what is happening today with tablets. – mins Nov 07 '15 at 00:02
  • "every pilot could customise positioning of controls and switches based on his/her way of piloting". This is sort of done by the Multi-function display, but it isn't about pilot customization. It's about saving space in small (ie. military) cockpits where space is at a premium. It also allows you to add new capabilities to the aircraft without having to redesign the cockpit. The downside is you have a bunch of similarly shaped buttons whose function changes with each mode. – Schwern Nov 08 '15 at 20:11
  • 4
    Have you ever tried using a large touch screen during violent turbulence? Try drawing a straight line on paper while sitting in a vibrating chair on max, only the paper is glass and even the slightest touch is a command issued – Tom J Nowell Nov 08 '15 at 21:32
  • Note that the Dragon 2 spaceship will in fact have mostly touchscreens for controls. However, the ship is almost fully autonomous, they are to be launched, controlled, docked (or birthed, I'm not sure), and landed roboticly much as the unmanned Dragon is. – dotancohen Nov 09 '15 at 12:22
  • The Russian Ka-62 helicopter has been shown equipped with a touch-screen 'glass cockpit'. Innovative technology might be more acceptable in the applications it is aimed at ("..cargo transportation, medevac and search and rescue operations, and can also be used in the oil and gas sector and for corporate purposes") than it would be in an airliner. – Spehro Pefhany Nov 09 '15 at 20:10
  • 3
    "every pilot could customise positioning of controls and switches based on his/her way of piloting" Wow, what could possibly go wrong? – Lightness Races in Orbit Nov 09 '15 at 23:48
  • 1
    @mins: I think we can do without spreading disgusting stereotypes about those with mental disadvantages, thank you. – Lightness Races in Orbit Nov 09 '15 at 23:49
  • Note: the Garmin G3000 which is mainly geared toward the smaller turbine side planes is already incorporating touch screens. – Dave Nov 10 '15 at 00:15
  • @mins: The well trained computer users use things like e.g. Vim, exactly for the reasons you mentioned: Keeping your hands on the keyboard where every movement carries so much more, and more precise, information. – DevSolar Nov 10 '15 at 13:47
  • OP have you ever tried to use toutchscreen keyboard? – Tomáš Zato Nov 11 '15 at 11:14
  • @LightnessRacesinOrbit Think of it another way: what could possibly go wrong if a switch is left in the wrong position? – Sanchises Nov 11 '15 at 14:25
  • @sanchises: I don't see the relevance of that. Misconfigurations can happen with either technology: that's not unique to either, and therefore not a selling point in either one's favour. – Lightness Races in Orbit Nov 12 '15 at 00:04
  • @LightnessRacesinOrbit Exactly my point - "what could possibly go wrong" goes for both options. Although indeed, I can only begin to imagine what happens when you accidentally load the wrong savefile... – Sanchises Nov 12 '15 at 09:06
  • I support iPads as electronic flight bags (don't ask, or do!). The problem I currently face is that the passcode requirement for email locks a device that is needed to reference charts, etc. on approach. I've got a solution, but I can imagine this is compounded for actual flight instruments. Tactile controls all the way. – blaughw Nov 14 '15 at 01:14

10 Answers10

118

Adding a pilot's perspective to what others have said:

Look at the switches and knobs in an airplane up close some time. You'll notice that nearly all of them have different textures,heights,sizes or shapes. That's intentional. Pilots train and train and train and train on checklists, especially the emergency ones. Muscle memory is a big part of quickly and correctly executing an emergency checklist.

As an example, flap levers/switches normally have a cap or top that is flat and parallel to the wings. It's easy to identify solely by touch.

Checklist item: "flaps up" Action: Hand to flap lever (automatic after doing it a hundred times in training), verify feel of lever, move, look to verify

Without the tactile element, every motion would have to be verified by looking before it's made. With the tactile element, the movement can be made and then verified visually while the hand is moving to the next item. In an emergency, every second counts and the savings from tactile + verify matters.

Beyond emergency procedures, they're simply safer. The plane I fly the most has a touch screen display and several non touch devices. In turbulence, the touch screen is basically useless. But, the important devices all have buttons and knobs. I can be banging my head on the ceiling (literally) and still adjust the auto-pilot or radio. In that kind of turbulence, the touch screen device is completely useless.

Sean.Mollet
  • 1,276
  • 1
  • 8
  • 3
  • 58
    This. 100% exactly! Try tuning the touch-screen radio in your car on a bumpy road. Now try it with a knob. No contest! – FreeMan Nov 06 '15 at 17:05
  • 3
    Ha, I just made the same comment above before scrolling down and reading this answer. I think the turbulence answer should be the #1 canonical answer; the rest of the answers here are all true, but the human factors seem like the primary reason physical controls persist. – egid Nov 06 '15 at 22:46
  • @egid perhaps I'm just interpreting it differently, but I'd consider turbulence an environmental factor, not a human factor (to me, "human factor" = "human error"). While humans are involved in the error that'd be introduced by turbulence, it's not directly their fault. – Doktor J Nov 06 '15 at 23:34
  • @DoktorJ Well, Human Factors as a concept is the study of the interaction between man & machine. Turbulence is environmental, but it affects how effective we are at manipulating the controls and is certainly something that guides cockpit/flight-deck design. – egid Nov 06 '15 at 23:36
  • 2
    Welcome Sean, nice first post! Hope to see more of those. – DeltaLima Nov 09 '15 at 21:54
58

Adding to the absolutely correct answer of @Mach1:

  • Flicking a switch gives clear haptic and audible feedback. Compare that to a touchscreen where you can never be sure if your intention has been interpreted correctly. This might not seem to matter much, but if you need to throw 20 switches in a hurry, the mechanical solution is head and shoulders above anything with a touch input.
  • Mechanical switches can be operated with gloves on. Touchscreens generally cannot.
  • If the computer driving the touchscreen crashes, you are out of control. Mechanical switches don't have this failure mode.
  • Mechanical switches are far easier to debug than touchscreens. You can make sure if the switch works by using a screwdriver and a voltmeter. And some knowledge of the routing, admittedly.

@Greg Taylor and @Lightsider are correct: Inadvertently flicking a mechanical switch is so much harder than one on a touchscreen, especially if it has a cage to protect it from movements in an off direction.

Toggle switch with cage

Toggle switch with cage (picture source)

Peter Kämpf
  • 231,832
  • 17
  • 588
  • 929
  • 4
    I'd add one small point, that inadvertently hitting a switch with a touchscreen would be far more likely. Brushing another area of the screen with a pinky or knuckle could change something you didn't intend. A physical switch doesn't change from such a light touch. – Greg Taylor Nov 06 '15 at 13:17
  • 2
    Especially if you're operating the aircraft in heavy turbulence! – Lightsider Nov 06 '15 at 13:43
  • 5
    @GregTaylor: Right, and for that reason important switches have cages around them so they need to be toggled by a movement in the right direction. – Peter Kämpf Nov 06 '15 at 14:03
  • 50
    As anyone who has driven a car for several years knows, once you are familiar enough with physical controls, you can keep your eyes on the road (or skies, or display, or altimeter, or whatever) and 100% use touch to find, check the current setting of, and change the setting of a physical control. Not so with a touchscreen control. – Todd Wilcox Nov 06 '15 at 14:23
  • 7
    I think the point @ToddWilcox makes deserves more visibility than a comment affords. I know it's a big reason for me to prefer physical controls in high-focus environments, even if those physical controls in the end change some setting in an electronic system (like the radio's volume level in my car, which uses a continually rotating knob to adjust the volume but provides clear tactile feedback that I am (a) touching the right control and (b) operating said control). – user Nov 06 '15 at 15:37
  • 1
    Touch-screens respond instantly when you touch them (except when they don't), while you can rest your finger/hand on a mechanical switch, prepared to activate it while not making any change. For example, the PF having his hand on the throttle levers on landing in case a go around is necessary. If that was a touch screen throttle - good luck. – FreeMan Nov 06 '15 at 17:07
  • 1
    @FreeMan I'm just imagining a touchscreen slider for the throttle suddenly jumping to TO/GA or idle because of spurious readings from the touch screen digitizer. shudders – reirab Nov 06 '15 at 17:52
  • 2
    There's also the fact that, with a mechanical switch, you can put your hand in the right, approximate area and find the switch by touch. You don't HAVE TO look at the switch. And if you're sufficiently familiar with the positions and what they mean, a simple touch will also tell you what position it's in without looking. Many of the mechanical switches require that you pull the switch outward before you can move it. So you're unlikely to inadvertently move them to a different position by touching them. – Meower68 Nov 06 '15 at 20:18
  • 2
    your third point is not really valid, people trust mechanical switches more, but I work on stuff like "the computer driving the touchscreen" and a lot more effort is put into making sure they don't crash than you would want. – Sam Nov 06 '15 at 20:31
  • 5
    Don't forget turbulence. Have fun tapping in a frequency on a touchscreen in mod tb; I finished a minute ago with knobs. Just imagine driving on a washboard or gravel road, trying to use a touchscreen that's an arm's length in front of you. – egid Nov 06 '15 at 22:45
  • @Sam You can put all the effort you want into making sure they don't crash... but any non-zero chance of failure is catastrophic if all your controls are in a single screen, and you can't tell me you can guarantee a zero chance of failure. Also, the operational environmental variables are infinite for all practical purposes, and designers can only test against and/or plan for a finite number of cases. – Doktor J Nov 06 '15 at 23:32
  • @Doktor J and you can say that mechanical linkages, electrical linkages and switches are test for all cases? – Sam Nov 07 '15 at 01:02
  • 3
    Further to @ToddWilcox's comment, when I got my current car I spent half an hour parked in the driveway practicing finding each control I might want to adjust while driving. If I cannot find and adjust a control by touch that control does not exist while I'm driving. – Patricia Shanahan Nov 07 '15 at 15:54
22

Mechanical switches gives a better visual inspection than digital touch screens. For a pilot surrounded with so many switches, just a visual glance would be enough to find the position of the control.

Mach1
  • 321
  • 1
  • 5
  • 1
    In addition to the vsual feedback, touching a knob without looking at it also provides a feedback on its position (I agree this is not enough to be sure the knob is in the correct position, but it is an information that can trigger some check process) – Manu H Nov 06 '15 at 14:23
  • 3
    As an additional point, when using a touchscreen your hand is blocking the view of it. You can't see the actual button that you are pushing, and you can't grab a switch, verify that it is correct, and then move it. – Adam Nov 06 '15 at 14:34
  • 4
    I wouldn't agree entirely with visual inspection. With a touchscreen, a good interaction designer can make visual inspection much easier. Besides, only the buttons which need to be shown at a time can be shown, thus reducing a LOT of visual clutter. – Nav Nov 07 '15 at 07:55
  • 2
    @Nav "only the buttons which need to be shown at a time can be shown" You mean, "only the buttons the interface designer assumed you'd need to be shown at a time can be shown." – David Richerby Nov 09 '15 at 08:53
  • 4
    @Nav "reducing a LOT of visual clutter" Is visual clutter in the cockpit a problem for properly trained pilots? I'll grant you it can be intimidating at first, but once you know the control layout, I tend to find "cluttered" panels easier to use in many situations because the control is always right there, in the same spot every time, no matter what else is going on. – user Nov 09 '15 at 15:16
15

Beyond the obvious aviation reasons of cost and so on, I can think of a few reasons why I'd not want an over reliance on touchscreens in the cockpit:

  1. If a touchscreen fails, you've lost all the controls that were related to it.
  2. Likewise, if a touchscreen (or portion of) fails then the whole thing has to be replaced rather than just the button
  3. Feedback - phones and similar devices offer things like haptic feedback to register a positive touch. I can't imagine that being implemented and working well in a fixed monitor. Pressing a button provides a positive reinforcement that you did press the button.
  4. Using touch-screens is hard if they're not still - it's far easier to press a physical button, without pressing an adjacent button, than it is to press an area on a screen without accidentally glancing another area.
  5. Some controls are designed to be easy to operate, while being hard to do accidentally - for example flaps, gears, mixture etc. All of these require some positive force.

With that said - you've posted some advantages and, in honesty, I suspect it'll all happen one day. So, for that reason, in the meantime I give you the stock answer to any "Why isn't..." aviation question:

  1. Cost
  2. Certification
  3. Proven reliability
  4. Market demand

Unless pilots and airliners are crying out for it, and a manufacturers think it will make the difference between whether they buy a particular aircraft or not, they simply aren't going to spend the time and money necessary to make it a reality.

Additionally, it'll almost certainly come in slowly. It's one thing to add some touch functionality to a flight computer but I wouldn't be expecting flap and gear levers to be touch sensitive anytime soon!

Dan
  • 9,308
  • 2
  • 33
  • 53
  • 2
    For new designs, I'd actually guess that the touch screen solution would cost less, especially since modern glass cockpits already have plenty touchscreens in place (and a lot of controls already have been assimilated into these.) Reasons 3-5 have a lot more to do with why lots of other switches still exist, though, imo. (And reason 6: having a touch screen control for the main power bus doesn't work out well, since it becomes very hard to turn it back on without power...) – reirab Nov 06 '15 at 17:56
  • 3
    "Proven reliability" A big deal in aviation. – T.J. Crowder Nov 06 '15 at 22:39
  • @reirab if you see a screen in a cockpit, don't jump to the conclusion that it is a touch screen. The downsides of a touch control do not carry over to displays so you might be noticing that instead of physical dials, a cockpit has multi-function displays. I would bet that what controls the display are traditional knobs / levers / buttons though. – Kelly S. French Jan 03 '18 at 17:56
  • 1
    @KellyS.French Yes, sorry, you're correct. That was a typo on my part. I meant to say "especially since modern glass cockpits already have plenty of screens in place." The point was that adding a touch-screen digitizer is not expensive (especially not relative to the cost of new aircraft.) Why touch screens aren't used for most of the controls has more to do with the other reasons listed here. – reirab Jan 03 '18 at 19:05
8

Touchscreen flight displays are already under production and the first commercial aircraft with touch screen primary flight displays as standard equipment to receive FAA certification is the Beechcraft King Air 250 turboprop equipped with Rockwell Collins Pro Line Fusion avionics.

There are some issues with touchscreen, such as,

...special security procedures are needed to avoid unintentional taps or movements that could jeopardise flight safety, especially during turbulence.

However, touch screens for aircraft are under intense development by almost all of the major avionics manufacturers and should be available for large aircraft by the end of the decade.

As an aside, the F-35 is the first combat aircraft to use a touch screen.

aeroalias
  • 100,255
  • 5
  • 278
  • 429
  • Very interesting. Would be nice to know what solutions are being explored to avoid unintentional taps. – Fabrizio Mazzoni Nov 06 '15 at 12:58
  • 4
    @FabrizioMazzoni: I've heard that taps are ignored, and large directional swipes (think iPhone's "Slide to Unlock") are the safe way to design for touch-screen controls in the cockpit. – Jacob Krall Nov 06 '15 at 14:14
7

"Wouldn't this make it more economic as there would be savings on material and wiring?"

Not necessarily, because the days of "one wire per function" are long over. Automotive has pushed that forward, and aerospace is following. Modern systems tend to be networked. This actually gives better reliability for less wiring, because it's easier to duplicate individual network cables than to duplicate a 6"-wide bundle of wires carrying the same information. Superficially the cockpit may still have 30-year-old switches and dials, but expecting 30-year-old wiring under the surface is generally incorrect.

"They could also be less susceptible to failures as traditional switches and knobs still rely on mechanical actions."

So do touchscreens to some extent. As for numbers of failures, how robust is your phone screen to minor impacts, compared to the switches on your car dash?

Graham
  • 2,404
  • 12
  • 15
  • "Automotive has pushed that forward, and aerospace os following." Really? Can-bus is older than Arinc 429? – Koyovis Jun 01 '17 at 20:25
5

Ive had touch devices misread my intentions many factors of factors times more often than mechanical switches. Furthermore you can operate mechanically arranged interfaces without looking, ie. by feel. In fact, having an interface that responds when you touch it is not at all what you want to happen when safety is a major concern.

Octopus
  • 149
  • 5
5

When my hands are swatting, or there is dust in the air etc I find that the touch screen on my phone does not always work. Also in turbulence, how well can someone touch a flat smooth screen in the correct location, but a physical 3d switch can still be used. Also what if the operator has to wrap his/her hand in a cloth due to it bleeding?

Do you wish your life to depend on the above?

However on the ground, a touch screen make sense for reading of event logs etc by a maintenance person.

Ian Ringrose
  • 439
  • 2
  • 12
2

The A380's maiden flight was in 2005.

iPhone's release was in 2008.

Planes development cycles are drastically longer than phones, and require very reliable materials.

We just didn't had readily available large capacitive touch screen when engineers started designing the A380's cockpit.

Antzi
  • 4,222
  • 4
  • 25
  • 46
  • 3
    But we did have resistive touchscreens, and the biggest advantage of capacitive is multi-touch, which is a whole extra layer of complexity. – Chris H Nov 06 '15 at 16:53
  • 1
    Indeed. But Resistive touch screen are drastically less user friendly, precise , and convenient to use. – Antzi Nov 06 '15 at 17:15
  • @ChrisH Capacitive touchscreens are truly touchscreens. Resistive touchscreens typically required some force: you had to actually press into the screen. Though that might actually be an advantage to prevent accidental touches... but it's also generally less accurate with a finger, and easier to accidentally break (thin plastic vs thick plastic/glass). – Bob Nov 06 '15 at 20:19
  • 2
    Nintendo DS are portable devices that use a resistive touchscreen along with a stylus. The device has very good response to touch, despite being resisitive. Also, I remember late-generation PDAs that had very good resisitive touchscreens too. – sleblanc Nov 07 '15 at 06:11
  • 4
    @Bob, I think you simply dealt with devices with low-quality resistive touchscreens. – sleblanc Nov 07 '15 at 06:11
  • Using a stylus to manipulate a screen is not very convenient ... – Antzi Nov 07 '15 at 13:50
  • @Sleblanc, My Sony Z Xperia Ultra is both resistive and capacitive (in addition to being water proof). It works very well, but spill a drop of water on it, or a drop of coffee, and it thinks the liquid is actually touching the screen. – Stephan Branczyk Nov 08 '15 at 05:12
2

Don't try at home experiment: Spill lots of water on a touch screen and try to use it.

Close your eyes and use your touch phone or tablet.

Try the same with knobs and switches.

Also, it is easier to specially and individually protect mechanical actuators against accidental manipulation than making hundreds of dedicated touch screens and protect those. It makes no sense if each knob/switch needs to be implemented as a dedicated touch screen, with a fallible control computer or some electronics behind it (in addition to the switching logic behind the knob or switch).

Arc
  • 121
  • 3