I’ve spent some portion of my life implementing touch screens for aircraft (usually single pilot). I’ve also sat in on more than a few PVI meetings that included pilots as we worked out flows. You can relax, at least a little (assuming the commercial folks use similar designs). The touchscreens on iBaubles and cars were designed by the art department. In the real world where actual utility is the goal, they’re much more usable and don’t look like a child’s coloring book. This will shock some people, but us research engineers discovered something that has evaded almost every single artsy fartsy designer in modern product development. There’s this thing called the English Language. And it has letters! And you put these letters next to a control to convey an exact meaning about its purpose. No really – it works perfectly, although some abbreviation is necessary due to real estate.
Unless Boeing decides to do something phenomenally stupid like contract their UI work to cheap programmers in Mumbai, the new touchscreens should be quite good.
I got JoseB to take me around on his Tesla and that was the one thing I didn’t like: too much of the controls is on a screen which is too large and too low to see it comfortably without taking your eyes off the road. Pretty neat otherwise; bit of a barge by European standards but that was to be expected.
I don’t really see the advantage of touch screens at all in a moving vehicle. I’d guess the idea is to squeeze more information into a smaller presentation area, and possibly increase the flexibility of the pilot interface. But even if you succeed at that, I can’t get away from the drawback of having to look at it rather than do some things by feel. And as I understand it, tactile feedback is big deal ergonomically.
There are also seems to be the inescapable increase in complexity that comes with any automated system (I think there’s an industry term for that?). So when you create flexibility by having a touch screen interface that can display multiple, multiple, multiple things to the pilot, you also create a more complex system for accessing those features. Even on an iPad that seems to be true, and you’re not trying to control a 500 mph vehicle with it (one hopes).
I’m not opposed to technological innovation, as a pilot or a person. I like a lot of the advances I’ve seen already in my career, particularly the various GPS-based tools we now have. I believe you when you say you’re doing things more intelligently than the marketing people you say are designing car interfaces. So I’m willing to be convinced - please tell me, what problem is actually being solved by putting touch screens into the next aircraft I might fly?
Having the option to change the screen to reflect what information is most important to you can be beneficial. The F-16 and 18 have MFDs that work quite well, but even they have all the critical (read things to keep the airplane in the air) set up for HOTAS (Hands On Throttle and Stick). Any vehicle that requires input from the human in the seat should be set up the same way, with physical controls that are ergonomic and intuitive. Another example is the computer mouse. Most people can at worst tolerate a trackball or mouse; but loathe touch-pads.
The question I would have is why two sister ships would have different control layouts. That’s a really bad design choice.
Differently-timed upgrades once in service and/or mid-contract change orders during fitting, resulting in the same nominal “class” having notably different systems, has been an aggravating issue in naval acquisition for a couple of lifetimes from what I read. Sometimes it’s minor and easily adaptable, sometimes it’s an operational headache. But they keep doing it because otherwise obsoloscence sets in.
It being such a schlep to approve and launch a new class of ANY platform (the US has been building versions of the Burke class destroyers since around 1988, and expects to keep procuring them another decade), there’s even more incentive to bolt on the new gear onto whichever unit you can get in the yard next.
The short version is that no two ships (or submarines) in the U.S. Navy are exactly alike. This is not usually much of an issue, because a sailor is assigned to a given ship for several years at a time. It’s not like a pilot who might fly one aircraft one day, and another aircraft the next day, and for whom consistency between aircraft is important.
Any sailor that is transferred to a new ship is required to learn and qualify on the new ship. Also note that while a sailor might be transferred from one Arleigh Burke-class destroyer to another, they are just as likely to be transferred to or from a cruiser, frigate, oil tanker, or amphibious assault ship.
The reason that ships in a given class differ is because the U.S. Navy continually makes design changes to newer ships and makes upgrades to older ships during overhauls. The USS John S. McCain (DDG-56) was ordered back in 1988, and commissioned in 1994. It was the sixth ship of the Arleigh Burke-class of guided missile destroyers, 67 of which have now been built out of 82 currently planned. With a class of destroyers that extends over multiple decades, you would expect newer ships to received upgraded electronics and computers, and for older ships to be upgraded over time. For operational and cost-control reasons, it is not feasible for all ships to be upgraded at the same time. So it’s no surprise that two ships of the same class might have different control layouts, depending on where they are in the overhaul/upgrade cycle.
I can explain at least one of the “whys”. Where I work, one of the products is switches, both the reliable clicky kind and capacitive touch. These assemblies end up being custom designs for each line of cars and there end up being reliability problems, especially with the touch style, simply because the design is custom and low volume.
Touch screens are made in immense quantities and you can differentiate between vehicles just by changing the GUI style and graphics. It allows for design reuse and because the screens are made in such huge volumes they end up being more reliable and cheaper.
But these are for things like adjusting preferences, pairing Bluetooth, adjusting exact HVAC settings - things you do not need to do in a moving vehicle.
Similarly, older airliners have thousands of tiny mechanical switches. Presumably most switches are not essential in the short term and many are rarely ever touched. So putting them in menus is an obvious improvement and it saves the aircraft manufacturer having to make these custom switch panels for each jet. Those have to be hand wired, volume is too small for automation. And the error and failure rate is going to be higher, especially in a system where there is more than one touch screen able to access the same jet system.
As for touch va mfd with side switches- touch advantage is you can make the screen itself bigger and can support more GUI styles.
i installed a new stereo head in my car a few months ago. I chose one that was NOT completely touchscreen. (It has a vertical space to one side with a volume knob and a multi-function button.) I could have bought one that was 100% touchscreen…and I (figuratively) get down on my knees and give thanks that I didn’t get the full touchscreen every time I use it. I’m forever scrolling too far, choosing the wrong song, or doing something else that I didn’t want to do because I’m reaching over with my arm extended and no place to rest my arm or wrist. I wish that there were MORE knobs (like a scroll knob) on the thing.
I feel exactly the same way about touchscreen tablets and mice. I carry my Surface 3 all the time and I always have a BT mouse with it. Touchscreens are fine…when you absolutely can’t use anything else.
Screens are great for displaying information but not that great for complicated/precise functions. I would definitely prefer mechanical/tactile controls.
Putting the controls of a vehicle on a touch screen is ludicrous, bumps, waves, turbulence or whatever and the operation actually becomes hit and miss.
A lot of system complexity can be reduced and made more intuitive if presented graphically. If you can pick destinations or standard routings by touching or dragging a spot on a screen, it makes more sense and is faster than trying to enter waypoint IDs via a clumsy keyboard (Ever used the G1000 twist-a-letter interface? Egads). If you decide to feed from “this” tank to “that” engine, selecting them by pressing their pictures on a graphic representation of the airframe & fuel system is easier. At least I think it’s easier than valve selections.
We have also designed systems with multiple views, or portals, which have a flight critical priority scheme built into the logic. If the pilot elects to concentrate on buried menus, we can keep a background attitude and reference display still apparent, and near his selections. We also have developed versions that had a fixed number of portals in the screen with attitude (EFIS) display defined as highest priority and the pilot is unable to remove or cover it. If another selection is made on the EFIS portal, it jumps to the next lowest in priority, always overriding displays of lesser importance, and will rearrange the others to ensure it’s as close to pilot LOS as possible. This is to lower the probability of getting heads down in a menu system and losing track of attitude and reference. Attitude info won’t go away and remains “nearby” (if that makes any sense).
Lastly, if an emergency or serious malfunction occurs, the system can present the screens necessary to deal with it immediately. Engine out? Graphic fuel system is popped up on the screen, pilot queried and presented with recommended crossfeed for his approval. Things that might be a few menus deep can pop to the surface when the system thinks they’re needed.
In the aircraft I am most familiar with, the crewstation is two MFDs with side buttons, and menu-driven displays. Systems that the parts are working within parameters are not displayed. Emergencies automatically reconfigure the displays to show the important functions.
The most often used controls are on the sticks. There are also the old school side console switch panels, but there are a lot fewer these days. Fire extinguishing is probably the most important of those, and it definitely isn’t controlled through the MFDs. Since the pilots always wear gloves I’m not sure you could use touchscreens anyway.
I take your points, but it also sounds a lot like what I mentioned - complexity going up as a result of efforts to bring it down. Sure, a good case can be made for the system automatically displaying certain information when an engine fails. But even there, the plane is now doing something I haven’t commanded and that can add to confusion.
The example of a graphic display rather than physical valves is interesting. I’ve read some on ergonomics that suggests many times it’s better to have a physical item to manipulate. In my current jet we do crossfeed / fuel transfer by turning one small rotary switch, and I think even that is an invitation to forget you’ve done it - indeed, I’ve seen it happen in this plane more than any other I’ve flown. I’d actually prefer to be moving a conspicuous valve or doing SOMETHING physical that makes the action more deliberate and memorable. Now, that can be addressed through procedures such as the Japanese “point and call system”. Maybe that kind of thing is going to be more necessary in the future.
Bringing this back to the article in the OP, I think we need to proceed very carefully going forward as we design interfaces. Technology is getting fancier and fancier, while humans are generally not. More and more, I’m feeling like I work for the automation* rather than the other way around and we need to change that.
Not so much in the plane, but I certainly feel that way when it comes to cars, phones, computers, etc.
My husband got a car radio/gps/who-knows-what-else with a touch screen and it’s totally impractical for the driver! Heck, even as a passenger, I sometimes have difficulties hitting the right spot on the screen to change radio stations or zoom out on the GPS. There is a volume button along one side, but it’s very low profile - I suppose if you drive it all the time, you get used to where it is.
I’m not anti-technology, but I am anti-dangerous, and IMHO, automobile touch screens are dangerous when driving.
When I first bought a car stereo that had a remote, I thought the remote would be useless to control a device that is within arms reach. But it was actually useful for scanning radio stations, especially on road trips where I wasn’t familiar with the stations. Just by knowing where the scan button was on the remote I could change stations without taking my eyes off the road. I would imagine that a full touchscreen device in a car might be easier to use if there was also a physical remote for common functions like volume control.
Touch screens first went into expensive cars because they were impressive high-tech, and gave you space for more options and more controls.
They’re going into cheap cars now because they are cheaper, more robust, and fill up the space used by the reversing-camera screen.
For ships and airplanes… well airplanes have been discussed above. Ships were /never/ easy to control, so it’s not a fair comparison to say “it would be easier to control using the old method”. The old method depended on having multiple trained people on the bridge: the Captain/Pilot/Officer didn’t even touch the controls.
From the discussion of some of the accidents, it semms like they are going to try to re-design the control system so that (1) Other people can see how the controls are set, (2) You’re less likely to make mistakes when untrained or sleepy.
All of those are good things. But looked at that way, getting rid of touch screens isn’t the main point
There’s another issue with touch screens which I hate, and which makes them dangerous:
Whenever and wherever you touch them, stuff happens. (duh!)
What I mean is that it’s too damn easy to make a mistake.
A brief touch in the wrong place, and you jump to a new menu, and it is not always obvious how to get back to what you want. And sometimes, it’s too late–you’ve already done damage; you just cancelled whatever function is currently running.
This happens on my ATM machine at the bank. Oops! I just brushed my glove across the “cancel” button when I wanted to press “other options”.
When flip-phones were common, and only a few people had smartphones, I used to pick things up with my fingers. (Shocking!But evolution gave us opposable thumbs for a reason.)
But I quickly discovered that if you pick up a smartphone the same way that you grab any other object on the planet, you’re likely to wrap your fingers around it. OOPS! touch the screen, while innocently handing a ringing smartphone to your boss, and see how you’v done damage, cutting off the call by accident.
I can easily understand why the military prefers , and needs, mechanical controls.
Knobs and switches are good!!!
and that was the one thing I didn’t like: too much of the controls is on a screen which is too large and too low to see it comfortably without taking your eyes off the road. Pretty neat otherwise; bit of a barge by European standards but that was to be expected.