Here is a side affect of having touch screens in new cars. As a blind person when my wife is driving I have to be very careful when changing the radio or music playing on Spotify as not to fuck things up or put her off driving. All so I cant change the AC in my half of the car. All so using my phone to play music is so laggy over blue tooth I cant use my screen reader to make things better. I all so have a friend who was born with one hand. He has to pull over if he wants to make changes to his radio or AC. In a car with nobs his stump can make most things work with out mutch of a struggle.
It seems that most marketers and automotive interior designers live in another universe. I really, really want to ask them if they really thought that this is appropriate and why.
And at least early on in the adoption of touch screens and probably somewhat still today the average consumer assumed touch screen was better because it was new.
Probably there is/was the "ooh shiny" factor of making everything be screens, and "ooh shiny!" sells/sold cars better than "ooh, functional!" which is also "ooh, looks outdated" in the mind of buyers of the last decade or so.
But after a while living with the stupid UX, the buyers would probably rather have the functional than the shiny.
"Now the screen is the button? Or is the screen not the button? I don't know! Maybe we'll all find out by the time when we're compacted like tuna fish in a can!"
Probably also to appear more modern. The problem with that approach is that, besides well known touch screens usability issues, it takes only one bump in the wrong spot to screw all the interface, so it's cheaper only until something goes wrong. They seem to know this though, as Space-X ships have all important functions replicated on real physical interfaces.
SpaceX did it on the Dragon too. Hard to imagine an environment that is more mission critical and less cost sensitive than a space ship under severe engine vibration or tumbling out of control though space.
I think that's different though. Spacecraft have orders of magnitude more features that need to be controlled than cars do. Just look at how busy the space shuttle cockpit was.
Hard to read in direct sunlight, too. (Partially-)Reflective displays with perfect dailight readability exist (used in aircraft), but they would likely not satisfy the cheaper-than-buttons requirement.
But cars without roofs are getting out of fashion anyway. Next step after no steering-wheel will be no windows.
This frustrates me to no end in our family car. Even if I "turn it off", which turns off the radio, the screen still shows a clock over a globe and still glows. There is no escaping it.
More generally: I want to be able to feel the state something is in. Whether it is push/pull, turn, or flip doesn't matter, but it has to physically maintain the state I put it in.
An exception to this of course are functions that are used temporarily, like the horn, and the wipers swipe-once. And if direction indicators turn off automatically, the input device has to change its state accordingly.
It is no co-incidence that the book "The Field Guide to Understanding 'Human Error'" By Sidney Dekker contains case studies mostly drawn from aircraft incidents. They're less numerous than car crashes, but more impactful and better understood.
I’m a flight sim nerd and it’s immediately clear when you have to keep track of more states than your are used to how much attention is needed and how much is freed up if you can reach over and feel the state of the switch.
I would love that for many functions in a car. And a knob with some bump or whatever at the ac temp setting, so I can adjust its position based on feel alone as well.
On a modern keyboard the caps-lock key would probably benefit from one too =)
I think switches that can set themselves to current state (flip back) are a bit pricey but can be had from $12 or so (iirc).
I completely agree with this.
I never understood how this became the new normal and mostly acceptable in the new generation vehicles (especially the EVs) in the name of progressive/futuristic design.
I hope the automobile manufacturers take a note of this, and bring back some of it that’s worked for decades. I believe few German manufacturers like the Mercedes-Benz to have a sense of this, hope they don’t get along with this trend too.
It's just cheaper to design and certificate one part and put it on multiple models than making dedicated analog controls for every new design. Touchscreens are just cost cutting measure. I wonder how many people died fiddling with car radios so one pencil pusher could get a bonus for skimming $10 off manufacturing cost.
Joking about the price aside I always loved that the original design goals for the F1 included that if someone found one in a barn in 50 years time they should be able to repair it.
I think that leads to quality, simple, tactile controls.
If a function can be controlled from more than one place (such as power windows or locks that can also be controlled from the driver’s seat or key fob, or anything controllable by software via central control panel) then you can’t really have a manual, state-full control for any of it since it can go out of sync with the state of the function.
The exception would be to put a tiny motor on each control to keep it synced up, but the cost to reliability (not to mention just money) would be unacceptably high.
As substitute for physical state inputs I could accept separate inputs for different states (this excludes off/on toggle buttons, but could be satisfied e.g. by separate off and on buttons) and a state indicator in a visible location, preferably close to the input.
Power windows are close to fulfilling this (up/down separated, and it might be argued that the state is sufficiently obvious, though questionable for the rear windows). A central lock controlled through a common lock/unlock button does not satisfy this.
Yeah and I guess some things, like door locks, are actual mechanical things that involve having a motor there anyway so it shouldn't be too hard to just expose some part of that mechanism in a way that indicates the state. This is done now with that little nubbin that sticks up, but I think can be made more obvious.
I think the big sticking point is usually climate control and radio since that's what people fiddle with most while driving. Maybe little things like the ride mode selector (but I'm not sure how many people actually fiddle with those sorts of things in practice). But nowadays the "radio" is rarely just radio or disc/cassette player. It's music, audiobooks, or podcasts off a phone. I don't really know how you do that with integrated controls that aren't screen based, like most luxury cars already have the little control wheel and also controls on the steering wheel. So what more would you need? Maybe some haptics on the control wheel to give you more feedback?
My car has a quiet indicator tick, it's so frustrating not being able to hear if the indicator is on. Particularly because they (Vauxhall/Opel) butchered the indicator stem switch meaning it doesn't click, which would be fine, except it returns to middle, which would be fine, except the dashboard indicator lights are hidden by the steering wheel, which would be fine ... if the indicator tick wasn't too quite.
Honestly it makes driving a real chore and is a huge safety problem; just that one little anti-feature ruins driving.
Couple of more such tests and modern car designers will maybe finally read the Design of Everyday Things. In this 1988 book the interface of cars was praised, but then the companies of course had to improve it.
I miss the buttons and dials from my old car. Muscle memory meant I could just turn on air con, change the fan speed, zone, volume etc with taking eyes off the road. Having to do this with the touch screen is bit of a pain in my new(er) car.
Touchscreens just seemed cool as they were a cool fresh futuristic tech and I were a kid. Now as I'm more rational and touchscreens are ubiquitous, it seems obvious to me analog controls are almost always better. The only exception is when the number of controls you need to fit really is too much so you better switch pages.
If they didn't do the test with people already quite familiar with the car's interface, I don't consider the results all that valid:
Most cars, most of the time, are driven by people who are very used to their particular interface. I couldn't find mention of this in the article, but it's necessary to state.
Furthermore, increased automation in many of these cars means you don't have to do many of the listed sequences as often, or at all:
"Lower the instrument lighting to the lowest level and turn off the center display." Why would I ever do such a thing when this is fully automated already?
My wife drove a 2002 Honda Insight for 2 years. She never did understand the "window override" switch on the dashboard that disables the power windows in the doors. She taped plastic over one side for a week the first time she hit that button and the windows stopped working. She took it to the mechanic a couple times "the windows stopped working again" because the switch on the dash was easy to frob by accident.
Why was that switch there in the first place? Being familiar with the car didn't help her remember it existed, even when she encountered the "windows wont move" problem repeatedly. Assuming you want a "stop the windows working" switch, why put it up in the corner of the dashboard and in a form where it's easy to hit accidentally and hard to tell that it's not in the right position?
The problem is deeper than screens and older than "legally required backup cameras;" the ergonomics of the car interior became secondary to "but marketing says people want feature $X" checklists.
Genuinely curious, how does she hit a button on the dashboard by accident? To me, the dashboard is behind the steering wheel and shielded from accidental presses.
Its up in the top left corner by the vent, and its a rocker switch. Quite easy to hit without noticing on entering or exiting the car, it got me several times too.
The switch typically disables the passenger and rear windows, and is there is that the driver can stop kids or pets controlling the windows. It’s been next to the window controls in every car I’ve ever driven that had electric windows.
To verify that a mountain was indeed being made out of a molehill, I checked the manual [1] for the 2002 Insight: page 78 explains this is the case in a single sentence.
I bet it would end up looking better as well. High quality switches and buttons feel really luxurious.
If you look at Gordon Murray's latest car, the t50, he spent a lot of time sourcing buttons and switches for the car because he cares deeply about their tactile feel. There is also not a single touch display in the car. Because he passionately hates them.
If you are into sports cars you kind of want to cry when you look at how badly some of the dashboards tend to age. Some of them look like embarrassing student projects where a 1990s web designer has crammed all manner of animated gifs with a horribly infantile palette onto the dashboard.
> High quality switches and buttons feel really luxurious.
That's part of why I have an Audi.
It's a 2021 model with a touchscreen. But it also has high-quality buttons and knobs for each of the tests in the article mentioned and would thus pass with a very low distance measured.
There are some weird issues, however. The audio buttons control the current audio source but don't fall back if the current audio source disappears. Like if you are playing music through your passenger's phone via CarPlay. If you drop them off at their house and keep going, the physical audio controls do nothing at all until you select a new audio source on the touchscreen.
I'd love to see similar testing conducted on "consumer" avionics boxes used in general aviation. Garmin has moved many of their units from fully-tactile buttons and knobs to touchscreens. This puts the onus on the pilot to fully master the system before entering a situation where you can't simply swipe/scroll your way to success. But using a touch screen in turbulence is nearly impossible; Garmin engineered a physical lip edge around the unit to hang onto to assist the pilot in stabilizing their finger.
There's a reason professional cockpits still largely eschew touchscreens when 250+ lives are at stake in the back.
I'd personally agree with this, but at the same time, I think aircraft like the F-35 have moved to touch displays instead of 4th gen MFDs (but works with gloves?), and didn't some of the SpaceX stuff have touchscreens. There must be some logic driving it in the defence/space industry...
I bet Lockheed Martin sells a $50K glove that works with the F-35 touchscreens.
It's not that bad, however. In a car you need to be constantly aware of the road ahead of you. On a plane, you are not required to have your hands on the controls while on autopilot and you can pay attention to the screens, as well as operate them - there won't be any wildlife crossing ahead of you, not any red lights forcing you to brake.
And if the plane sees a threat it will warn you well before your human senses can so you can pay full attention to it (and the helmet-mounted displays). I assume the F-35 also has stick-mounted controls to operate in the helmet display.
Touchscreens are potentially very useful in a fighter (maps, general SA, designation etc), but not for everything and not in every situation. But just because you get a touchscreen doesn't mean you have to forfeit frame buttons with onscreen labels, HOTAS etc.
Which should kind of be a sign that for any important high-stakes task you should be using physical switches.
I predict in 10 years car manufacturers will bring back physical controls for some things (if not all).
One advantage physical controls have is that I can operate them sightless once I learn the layout. Most touchscreens I've seen in cars don't really have that feature because of the design of the system behind it (whole screens shift so returning to navigation isn't often simple, for example).
> I predict in 10 years car manufacturers will bring back physical controls for some things (if not all).
Maybe. If new cars are all self driving, I could see manufacturers keeping with touchscreens. It's safer to stay distracted longer if you're not in direct control of the vehicle.
Touchscreens are fine for performing car-related settings and tweaks.
Eventually if we get to actual full-self-driving they might go that route. That's way more than 10 years out though, I think. You're talking about a point where cars are able to not have a driver at all before that becomes a reasonable option IMO. That said, even elevators have buttons.
Exactly. Why can't we invent better manual controls that don't wear out as quickly, rather than switching entirely to something as problematic as a screen?
Or just design the manual controls to be easily replaceable.
Not hard to make an insertable switch component that snaps into a slot, and which can be removed and replaced in 10 seconds using a simple tool (or maybe no tool at all).
Same general trend on marine MFDs (basically the equivalent thing for boats). Models with knobs and buttons are the "premium" version, the default is typically just a giant touchscreen, which is not always great when you have wet hands.
The flat screens look good, and work great in calm situations, but in heavy seas and/or rain, they can be challenging at times. In my case the manufacturer (Raymarine) offers a wired remote control, so I have a knobs-n-buttons controller that is easy to reach and offers more direct control when needed.
There is one practical argument that can be made against mechanical switches and buttons, and that is they will eventually fail from long-term use.
I've heard from friends in the aviation industry that pilots take extra care to put as little stress on switches and buttons as possible during normal use to prolong their service live.
The uninitiated might think why bother when a switch or button is dirt cheap, like several cents per unit cheap. And they would be right, the best kind of right. But when a switch/button does inevitably fail and needs to be replaced, the cost can easily come out to at least several hundred bucks between the labor, reinspections, and recertifications among other red tape that help ensure safety.
So if (if!) touchscreen interfaces are more durable and last longer, that is one fair argument in favor of them over mechanical switches and buttons.
> There is one practical argument that can be made against mechanical switches and buttons, and that is they will eventually fail from long-term use.
Before a touch interface fails from long-term use? I highly doubt it. Plus, a switch is a very easy component to replace. Touch screens can be but aren't always.
I think service life arguments are just poor effort. We well understand the appropriate average service life of a variety of switches. We don't understand the same for touchscreens, especially modern ones, as they haven't been around as long.
Also the touchscreen is more of a single point of failure. If the touchscreen fails, a whole range of systems and functions are affected. If on the other hand a single switch fails, only one function/system will be affected.
In my twenty years of driving I have had a single manual control fail in a vehicle and it cost less than $50 to get it fixed. I can’t imagine fixing anything on these modern cars being less than several hundred.
Do keep in mind I was talking about replacing failed switches in the context of commercial airliners. As far as ordinary cars are concerned, I agree those would be dirt cheap and easy affairs.
My experience after flying 737, MD80, A320, A330 and A340 is that nobody takes special care with cockpit buttons. They are work tools and treated as such.
The main ones (autopilot and flight controls), rarely fail if at all.
The only ones failing from time to time are small switches for radio channel volume or cockpit light adjustments and system buttons at the overhead panel that are easy to replace by maintenance.
Touchscreens are not a good option for main controls due to poor visibility(dirt from fingers and sun reflections), hidden submenus, turbulence making hard to press the correct button…
The A350 and 787 are using trackball controlls for submenus and the onboard computers, not a touchscreen.
I must add that one place where we are using touchscreens is in the fly documentation. Most airlines use somekind of tablet, ipad or surface with apps for performance calculation, navigation charts, pdf manuals, etc… they are working mostly ok now a days and I’dont think you can substitute the touch screen with buttons for that without loosing a lot of functions.
It is for this reason that I have almost no Garmin avionics. I use Avidyne IFD GPS navigators which can do almost everything without using the touch screen. Even then I find the knobs are too easily turned to use very accurately in turbulence. To use them or the physical buttons in turbulence it is necessary to brace part of your hand or some fingers on the bezel, which is fine.
Critical checklist items for a smoke filled cockpit have zero to do with avionics. And those items are practiced so much that any pilot should be able to handle smoke in cockpit with their eyes literally closed.
One could argue that the EGT and other secondary engine items could be hidden on secondary tabs on something like a G3X. Unless they own their own aircraft, I don't think most GA pilots can do much with their eyes closed.
>using a touch screen in turbulence is nearly impossible
This is one of the biggest reasons I dislike touch screens in cars, yeah. Tons of roads are more than turbulent enough to make it hard to hit buttons. Not having a physical edge / clicking / etc to tell you where you are and when you've done a thing means you have to use your eyes, which means disabling what is by far your biggest safety tool while driving.
IMHO touchscreens in cars are the beginning (or perhaps a continuation) of the crapification and tiktok-ization of this entire product category.
I remember once driving, I think a Renault Megane, which had all the controls (stereo, climate, etc) replicated as buttons on (behind) the steering wheel. A slight learning curve, but completely seamless driver-car integration once learned. Someone had obviously thought carefully about how this should work, instead of just slapping in a touchscreen and a bunch of menus.
When I read your comment, I immediately thought of a car which you exit by removing the floor plating, after which the car lifts itself up, so you can comfortably step away from the underside. Can't wait to see my grandma try that!
Oh hey, this reminds me of a car I saw. Thankfully did not have to ride it. The floor on passenger side was missing and it was a bit lifted , so you could get out the car that way. Was handy because sometimes the door lock got jammed and refused to work. Same goes for the meatgrinder-style window controls.
Other than that - it worked and somehow had all the needed papers to be street-legal.
The "crapification" you describe is the creep away from the scientific
principles that once underpinned this field. Before UX we had HCI
(Human Computer Interaction) which was in turn a development of CE
(Cognitive Ergonomics) and other "human factors" sciences.
These sciences were rooted in very rigorous but time consuming tests,
observation, psychology and physiology.
from TFA: "Designers want a "clean" interior with minimal switchgear"
This is where the wheels fall off the wagon. Should "what designers
want" be high amongst the priorities for safety critical products?
Genuinely curious, what do you like about it compared to Win10?
I've only tried it in a VM for a few minutes so far, but was unnerved by the general feeling of 'pretty, but impractical', mainly thanks to the taskbar and the right-click 'hide everything by default' context menu.
Evidently not if the work they're producing is reportedly outperformed by old school physical controls from more than a decade ago and in most of the vehicles tested it wasn't even close.
Often they recruit kids with graphic arts backgrounds, hand them some fancy post-it notes and a YouTube video of how Zipcar did a journey map, and set them loose.
UX usually focuses on the critical path for the top-5 tasks. So turning on the car radio makes sense, but changing the radio station didn’t make the cut, so some rando engineer guy stuffed it in a menu.
When it’s done well with a great team and time it’s magic. It’s easiest to see when Apple gets software right, like Keynote - the functions of making a presentation are immediately obvious to an elementary school student. But even then, once you leave the happy path, woe to you - modifying a template is a dark art to most people.
Or you could use Apple's iTunes as an example of how to build one of the world's worst and most user-hostile interfaces, but one that every iPhone user must deal with unless they let Apple have complete access to all their information via iCloud.
I'm convinced most people really don't like iCloud, but since the alternative is iTunes, they basically have no choice...
No it's not their job, and I'll try to explain why I think that.
Apart from the remit being just too broad, designers in any case are
part of a complex team that deal with a multitude of functional,
non-functional, regulatory and financial requirements.
Now, we have many different definitions of "designer", which I am very
aware of, but I believe that, in some circles "designer" has become
romanticised and extended to include a set of perceived "magical"
powers to "deliver what a boss wants". That is a distortion of the
role to something grotesque.
Speaking from a domain in which I have expertise; in sound design a
great battle ensued between designers, users (audiences) and the
'bosses' (studios and publishers) as to how music and films should
sound. You probably know this as the "Loudness Wars". I think it
remains a textbook example of misalignment between technical, artistic
and financial factors. It also remains an example of why I think
"Markets are a myth" [2].
Despite listeners saying over and over that they "Don't want it", the
producers, through a mess of internal motives (mainly financial),
repeatedly foisted their values onto them, being obsessed with what
they think users want in preference to flat-out contradiction that
would be evident in even the most cursory market research.
The job of a designer is to balance factors, and in a sense act as an
advocate (stand-in) for the user by mentalising their actual
needs. It's a very demanding and complex skill. Doing "what your boss
says" is absolutely not it and reduces a designer to a tool.
On the other hand, a job of the designer is also to listen to expert
technical advice outside of their skill-set, and so must not get
carried away with any grand "aesthetic vision", wanting to be Steve
Jobs.
A hard line to tread, and one requiring strong will and ethics as well
as judgement.
Related, the challenges sound designers face in making dialogue audible. What seem like simple problems (make car climate control buttons easy to use, make the speech in a movie easy to understand) turn out to be incredibly complex.
I worked briefly as a freelance experience designer hired by an appliance manufacturer. I asked if they could send me physical prototypes of controls so they could be tested. They refused and said it would be too expensive. They expected the controls to be designed, spec'd, and sent to the factory without any usability testing.
Designers can do all those things, but often they're not given the space to.
The best products are typically produced in an environment where the people running the company care about the design. This is a rare environment.
> from TFA: "Designers want a "clean" interior with minimal switchgear"
Speaking as someone with an Interaction Design (IxD) degree: no we fucking don't. Tactile buttons being superior has been known for ages. For example, Bret Victor wrote "A Brief Rant On The Future Of Interaction Design" in 2011, so over a decade ago[0]. Not that anyone with the power to change things listened, because these decisions aren't made by the designers.
This is mostly a consequence of people higher up trying to save costs by using touchscreens, which is cheaper to buy and cheaper to develop for. HCI and IxD have always had this issue that we're asked to fix things up after everything else has already been decided. Basically, we're mistaken for graphic designers who decide on what the final product will look like. So we're given a touchscreen to develop an interface for, not a blank-slate car interior (or whatever) for which we get to decide the button layout.
At the risk of pulling a "no true Scotsman", this is a consequence of cost-cutting first and foremost. Don't blame the people who actually have a background HCI or Interaction Design. We all knew this was coming, and we hated it. If we're told to make do with the touchscreens we are given, with the alternatives of actual physical buttons being ignored before we even get to make decisions, then don't blame us for the lack of those buttons.
The cost of fixing hardware failure in a final product is not the same thing as the cost of developing and mass-producing the product.
For example, we're not talking about one button, we're talking about a lot of buttons, usually custom-made for the car in question. The whole dashboard physically has to be designed around them. Meanwhile Tesla just slaps a screen on a mount in the middle of the car and calls it a day. It's basically "we have to get everything right the first time" vs "fuck it, we can always fix things in a later software update". Which is a way to save costs by cutting corners.
The buttons all have their own complicated logic too, although I suppose that even with physical buttons one can handle almost all of that purely through software these days, so that's not really as much of an issue any more as it used to be (it does make me terrified that cars can be hacked and bricked, but I digress).
Speaking of a lot of buttons, that's the other thing: if all your buttons are virtual, you can have infinite buttons! The only thing we have to do is introduce a ton of mode switches! Which is absolutely terrible when you're driving, but nobody seems to care! So we can cram a ton of features into a screen that would otherwise require a million buttons, and use that in marketing. Even though we'd probably be better off if some time was spent to whittle things down to the essentials and design the interface around those cleanly.
Many buttons also means many more pieces to physically install, and many many more wires. And each one (or small cluster) is often accompanied by even more independently-wired small information displays (small LCDs and LEDs for showing the state / temperature / etc) which are yet more wires.
A touchscreen is largely just a single fused physical unit with ~two cables: a data ribbon and power. Utterly trivial to install and wire up in comparison.
Most executives don't have the vision or creativity to come up with these trends; they have to pick them up from somewhere. I think there's plenty of blame to go around though.
The trends adapt to the requirements of the customer. See also: the appeaiance desktop interfaces having phone interfaces that don't fit desktop affordances at all.
Note that "customer" can be a manager or similar higher up in the hierarchy.
Bret Victor article is very good. Thanks for sharing.
Some remarks stood out for me:
> talk about technology. That's the easy part, in a sense, because we
control it. (my emphasis)
Yes, I agree with him strongly. But - there's been a dreadful
anti-intellectual tide this past decade - a descent into
"technological determinism", or the idea that technology is its own
process to which humans must bend. It's the idea that we don't
control it. It comes along with the overuse of words like
"inevitable", "ubiquitous", "unavoidable" and endless talk of cats
escaping from bags and genies refusing to go back into bottles. It's
a defeatist and lazy creed that seeks to excuse a race to the bottom
of cheapness, as you describe, with a narrative about how we "have no
choice".
> if a tool isn't designed to be used by a person, it can't be a very
good tool, right?
Increasingly, tools are designed to be used by other tools. Humans are
being sidelined amidst the interplay between machines. For example;
the demise of the Web is largely due to bots and the arms race to
create other gatekeeping bots to defeat them.
> Hands
Bravo! Not "a finger" or "your thumbs". That's why I use a keyboard,
interact through text-based technology, and cannot fathom
thumb-twitching smartphone users. I totally get what he's saying,
having worked in sonic interaction design with musical instruments
(NIME) stuff like the ROLI seaboard (or whatever they changed the name
to)... hands and touch, with mechanical haptic feedback is the way to
go.
I wish more people payed attention to this understanding of our
relation to technology as embodied beings, instead of chasing a
"clean" disembodied dream - which I think hides within sublimated
Orthodox Dualism in the tech community - but that's another story.
I have a bachelor's and master's in Industrial Design. When I first entered the software industry after grad school in 2000 a master's was the floor for work in UI Design or Information Architecture (Ux wasn't a job title at this time). Many of the people I worked with in these early days were CogPsy PhDs. Design was slow and methodical. This seemed to hold true for the next decade or so. As design as a competitive advantage (or necessity) started to take hold more and more people flocked to Ux. Many in the field today are self-taught, attended bootcamps, or pivoted away from graphic design (thanks Dribble) to Ux. Did we lose something when many Ux practitioners no longer have roots in HCI, library sciences, human-computer interaction, industrial design, human factors? I'm not going to judge. Myself, I transitioned from Ux to programming.
This started years ago, my example is the transformation of the Jeep CJ into the wrangler. CJ were fairly cheap, worked fine and came with no frills (I won't talk about rust). The wrangler is an overglorified SUV. When that happened, I went to regular automobiles, which are now getting harder to find these days.
I can understand where you're coming from, but compared to the vehicles I own now my 98 TJ seems "no frills".
The soft top folded down pretty nicely. The rear seat could also fold down, as did the windshield. The half doors were still removable, and there were still only 2 of them.
It was a manual transmission but had A/C. The heater had two settings: "lava" and "off" :D
My cars heater started bellowing hot steam into my car through the vent system a few weeks ago. I took it to a friend who is mechanic and asked how much it was to fix it. He said "we live in the tropics, what would you use a heater for?", then disconnected it completely and charged me a beer.
I see the analogy. TikTok has streamlined the consumption UI to the point where your control is extremely rudimentary -- you like it or don't. The car makers are doing the same, to some degree -- you follow the 95% use case or you experience pain. I had a rental car last weekend for 4 days. I turned on the radio and didn't like the station. I tried to figure out how to tune a different station and couldn't, so I just turned it off. Perhaps they should switch to a model where they just play me random music and I say "yes" or "no" until they learn how to train me what to like.
Yes, and just wait until they turn on ads. You know it's coming. 'Two for one cupcakes this hour only at Let them Eat Cake!, coming up on our left in 30 seconds. Pull over Yes/No?'. Etc.
Gosh darn KIDS these days TICK TOKING all the DARN time, back in MY day we'd just PLANK on RANDOM THINGS then hit the WHIP and yell YEET and post it on INSTAGRAM for LIKES.
My car radio is completely unusable. I had it for 2 years and never figured out how to tell it to just play everything on the usb. Instead it groups by artist or genre and only plays that.
Eventually the usb port just stopped working so I use VLC on the phone and stream to the car via bluetooth.
I wish I knew what the designers/engineers were thinking when they created the ability to play music from a USB drive. It feels like their target audience is not the type of person that would put music on a USB drive for the car.
I just want it to recognize that I sort my music into folders and show the folder hierarchy. Instead, in my last car, it would flatten the entire structure and then sort by filename. To top off the shitshow, it would also only show up to 99 files, and scrolling through the list was painful. Each tap on the scroll button would only move one line, not one page, so if I wanted to play the 80th song, I'd have to tap to scroll 80 times.
seriously though, the USB audio player is a dumpster fire. They completely redid it with the last update -- and it got worse, not better.
some random examples off the top of my head:
- switch to another input source and you lose your audiobook position
- sorting is completely utterly broken
- audiobook has no skip-back-a-little function
- soooo many bugs
The only nice things I could say is that it mounts linux filesystems plus it plays FLAC and Apple lossless directly.
I learned that when the touchscreen starts flickering, it means the alternator has shorted out and you've got only a few miles before the car goes dead.
Really, these people are poison to engineering culture, and once that deteriorates, the company soon follows. Look at Palm, HP, Intel, I'm sure there are numerous others.
I think Apple's design choices provide a better analogy but the point is still taken and I tend to agree. They want cars to be disposable technology that consumers are continuously upgrading (like their phone)
One change is that a car ceases to be an 'object' (to show off), or a 'tool' (to get you to work), and becomes an 'experience'; of course, controlled by corporations.
I really dislike touch buttons in car, for some reason looking at that screen is a distraction that feeling for a button is not. Even worse, though, are physical buttons that digitally change. I drove a Landrover like that and it was very, very frustrating.
I hardly know anyone who uses an internal GPS provided by the car. Everyone uses their phone, either through CarPlay or just putting the physical phone up somewhere, somehow. Every car manufacturers' GPS's I've seen or heard of have terrible UX compared to phone apps such as Google Maps, Waze, or Apple Maps.
Polestar, and probably new cars from the sister Volvo,have a really good Android Auto based (not the screen cast Android Automotive) navigation screen using Google Maps for navigation. On electric cars it uses information from the car to build better routes, estimate battery usage and suggest charging stations.
I'm so happy to have physical buttons on the steering wheel because the touchscreen, powered by some obviously ancient chip(feels 2009-ish), is mostly useless, aside from maybe showing the map.
That being said I don't actually use any features during driving, except for adjusting climate control and compulsively checking fuel economy.
I'm curious what features people miss now that physical controls are mostly gone.
Touch screens are just their 60+ year old fossils deciding "that's hip, that's what kids want!" and probably their testing audiences responding more positively to images of flashy touch screens and shiny lights.
Driving a car with touch screens (new BMW or Mercedes) has left me very unimpressed. My 2016 VW Golf has actual buttons, switches, and knobs to twist and turn and press and flip.
Car reviewers, too, often say it's a shame that car manufacturers are switching to touch screen nonsense. It's such a shameful trend if you think about it. The BMW series of pre-2022 had buttons in the dashboard, but the upcoming new series will do away with those entirely.
Touch screens even find their way onto steering wheels and doors.
Of course, it's easy to understand why:
1. It's cheaper to produce;
2. It looks more expensive, so the price goes up;
3. Testing audiences respond positively to shiny lights;
4. Fossils decided that this is what the young people want.
Honestly, I hope European legislation makes it illegal at some point. For the sake of safety. With touch screens, even the most simple task requires you to take your eyes off the road in front of you; with regular buttons you could do many task just with touch.
What was even more surprising, to me, is that Mercedes had this amazing nice center console unit to control things with your arm in a rested position. They removed that piece of brilliance!
So, now you need to do everything with an outstretched arm in a moving vehicle to operate tiny buttons on a flat touch screen.
Oh, and the touch screen can only barely hit 60 frames per second and often feels much slower. They're even saving costs on GPU power in their fancy luxury cars.
> Touch screens are just their 60+ year old fossils deciding "that's hip, that's what kids want!" and probably their testing audiences responding more positively to images of flashy touch screens and shiny lights.
It's mostly a cost saving measure.
Physical buttons are expensive. I you eliminate them, the car gets cheaper to make.
That's all.
It's a sign of low quality and I expect that in 5 to 10 years, consumers will start to realise this.
I guess its the installation. One touch screen is one process. 15 buttons require 15 operations to finish the dashboard.
You have to save everywhere, else costs will run up. Small savings become huge at scale.
104 key mechanical keyboards have 104 buttons and may go for $30. Maybe it is the knobs? But you can get a midi keyboard with 8 knobs, 49 keys, 12 pressure sensitive pads for ~$100 (but maybe it wouldn't last in car cabin heat).
Those keyboards don't have the same environmental requirements. Automotive environmental is hard. Temperature extremes, high and low, plus humidity extremes.
There are also reliability expectations. If I need defrost because the windshield just fogged over, it better activate when I turn that knob on my sixteen year old car. And so far, it always has.
Years ago when I was working in IT at FoMoCo I recall seeing a piece of paper on an office bulletin board outlining how they had managed to save like $40 on the production cost of a Taurus, a vehicle that at the time was about $20,000. Those savings were the result of multiple sub-$1 to several dollar cost savings tweaks made between production years.
Ford has built something like 8 million Tauri, save a few dollars on each of them and it adds up to real money, like enough to redecorate the executive cafeteria.
Every little counts when it comes to exec bonus time.
My favourite feature of touchscreens in cars is when you try to click a button but go over a bump[1] so your finger misses and you press something else. Genius.
I do get that touchscreens allow manufactures to add and remove controls though.
Rolling out UI changes for self-driving cars, like getting rid of the, knob behind the wheel, will help with safety no end.
1. Not sure what the bump was, probably the neighbour's kid or dog or something. Too busy trying to get the latest Smartless. That Will Arnet, what a card, etc, etc...
The reason cars are as insanely cheap as they are for the level of manufacturing and design sophistication within them is because of many, many such cost reductions across the whole vehicle (which is made possible by the large scale on which they are manufactured).
> It's a sign of low quality and I expect that in 5 to 10 years, consumers will start to realise this.
Consumers is the key word here. Manufacturers already know that buttons have the "disadvantage" that they break independently. They are also easier to fix with a generic replacement part. Which means that you won't have to scrap your car because it suddenly became unusable.
Force manufacturers to provide replacement parts for 25 years after original purchase, and see them flocking back to the basics. But that prob. won't happen in EU (because it's against the interests of Germany) or USA (because "communism"). So I guess we're depending on the common sense of Japanese and Korean manufacturers?
> It's a sign of low quality and I expect that in 5 to 10 years, consumers will start to realise this.
I'm not so hopeful. The same can be said about household appliances. Yet more and more random things figure they should have touchscreens, or at the very least touch buttons. And this trend has lasted for far more than 10 years.
The impressive thing about the touchscreens in BMWs is that you don’t need to touch them at all. [I’ve not seen the new models though. I’m sure they are not better than the ones being replaced.]
> The main part of the BMW iDrive system is a control wheel, which can turn clockwise and anticlockwise like a volume dial. It can also be pushed forwards, backwards and to each side as if it were a joystick, and the centre acts as a button that can be pressed to confirm a choice or select an option. As mentioned above, later versions have adopted touchscreen technology, gesture control and voice commands, so there are multiple ways to operate a newer iDrive system in addition to the rotary control.
It's a physical user interface with buttons and a joystick. Which is basically what everybody here wants.
5. The manufacturer can change functionality and user interface with a simple software update.
If Toyota half-way through shipping their latest car realize it's better to have two knobs on the dashboard, they can very easily add one if the dashboard is just one big touch screen.
In some way that's even worse though. Everything should stay put. When driving a car, all this stuff is a secondary activity. I need to be able to develop muscle memory to ideally perform these finds blind while giving my main attention to the road. Buttons help doing this without looking. I might be able to do navigate a touch screen quickly if everything is in the same place all the time. Moving things around is just another opportunity to force more attention to the secondary activity
What this "we can change it later" option creates is designs that are not very well thought out. When you have physical buttons you must be double sure that this is the best layout that you can come up. You need to commit and double (triple) check with multiple people. Allocate resources for manufacturing/tooling. But you are forced to think about it really really hard.
With touch screen and OTA updates, you can skip the hard part and leave it for future you to improve if needed. But as we all know, when it's already sold there is no motivation to spend money to improve. So touch UI stays half baked. And only gets improved with future models.
I think the trend towards touchscreens has to do with the halo effect of the iphone. The fully touchscreen phone was much more modern-feeling, and also better and easier to use than previous phones with buttons.
The irony is of course that the decision to have very few buttons (not one, not zero, but very few) with almost all input via the screen was made very carefully by Apple with very specific justification based on understanding of how phones were used and could be used. This is clear from Jobs' iphone keynote.
If Steve Jobs, Jony Ives etc were redesigning car interfaces it's far from obvious that they would have made similar decisions.
Physical buttons: 1. Are expensive 2. Need space on the PCB 3. Need ICT 4. Need special soldering sometimes 5. need a dedicated interrupt interface on the microcontroller ( that's why are more responsive) 6. Need software both at "kernel" (BSW) level and at userspace (application) level. A touchscreen "button" needs only a callback to a routine and a lot of patience from the user.
There is one benefit - you can update the UI of a car with a touchscreen but not one with buttons. Tesla's first touchscreen [1] now looks slightly dated, but they're able to just update the entire fleet.
No doubt it'll get to the point where you can't update it any more - either due to hardware incompatibility, lack of processing power, or some new technology being added. But it has meant that a 2013 Model S looks more modern today than it would have otherwise.
Equally, tech tends to look dated much faster than physical buttons do. It's too early to really say which has more long-lasting appeal.
The change is important during product development, though. I think touch screens are a reaction to PMs at car companies wanting to make last-minute changes, and being told "no, we already spent 10 million dollars on the injection molds". Put it in software, then your lazy engineers just have to stay late for a month. And if they don't finish in time, hey, just fake it in Photoshop for the ads and update the UI later!
I feel like the contribution of project managers to the humanity is net negative. UIs, whether on screens or as hardware controls, need to be built to suit the human body and to not require thinking to operate once one develops muscle memory. Every other concern — including aesthetics — is secondary. Touchscreens in cars are very contrary to that because they require visual feedback.
>you can update the UI of a car with a touchscreen but not one with buttons
A car is a dependable tool. Changing the UI during a car's lifetime is dangerous and unprofessional. I'd say the same is true for smartphones and computers but I guess the majority of people think of them as simple "cool entertainment devices"
Speak for yourself. My car gained the ability to display directions from CarPlay in the heads-up display overnight, while parked in my garage, increasing the value to me massively.
There is one benefit - you can update the UI of a car with a touchscreen but not one with buttons.
This is a bug, not a feature.
If I'm driving then I'm driving. I want any non-driving controls to be as simple, consistent and reliable as possible. I don't want any non-essential controls at all. I don't want anything I might want to use while driving that requires me to take my eyes off the road at all. I couldn't care less what some flashy touchscreen UI looks like because I should never have to look at it.
The physical controls on the dash of every vehicle I drive regularly still work as well and feel as comfortable to use as they ever did. In some cases those vehicles are over a decade old. I'll take that over the modern touchscreen junk any day.
It's interesting to see a parade of people object to the updatability.
Sure, on a minute-to-minute timescale, anyone must obviously agree.
But over the long term of owning a car, it is an immensely valuable feature. My 2018 car still feels quite new and fresh - much less reason to replace it than if it were falling behind.
Not with any urgency of course; it's one factor among many. We have a 2015 car that feels like 2005, and a 2018 car that feels like 2022. The terrible map/etc. experience will certainly be a motivator when replacing - and makes old car worth much less on the used market than one that updates.
> You'd buy a new car because the UI feels a bit dated?
There's a substantial group of people who lease cars, and just get a new car after the lease is up. For that group of people, that's probably a substantial reason along with the exterior styling.
It absolutely is not. I do not want my 2 ton moving vehicle updated on a whim. My 2019 has a screen I can never really turn off that is bright at night, my 2008 has a slow laggy UI that an update will never fix. The 2000 Miata sitting in my garage has the best interface of them all. Push buttons and dials for climate control, two window switches in the middle tombstone area and that's really it outside the typical steering wheel controls for signals, lights and windshield wipers. It's amazingly simple and should continue working even when my newer cars are dead and gone.
For devices that connect to the internet, such as for updating maps and real-time traffic data, does the lack of security updates mean that your 2-ton moving vehicle is now somebody else's 2-ton moving vehicle?
Completely agreed. Mainly, I wanted to make a distinction between "cannot be updated" and "can be updated but isn't". The former is a design choice that removes entire classes of vulnerabilities. The latter is a usage choice that keeps vulnerabilities open.
Since I cannot directly reply to you (ajconway) -- this is thoroughly untrue. Tesla does it at the very least and I'm aware that is becoming a "feature" offered in other vehicles now too.
If you click/tap on the time stamp next to a message you reach a dedicated page where you can reply (same thing I had to do to reply to yours at this conversation depth).
I'm with you on this, I do understand people not wanting the extra cognitive load to learn new changes in the UX but as a tinkerer I really love that my car can get OTA updates that add/changes features.
Actually many (most of?) Tesla owners have a Tesla also for this reason. Source: lurking in Tesla owners forums/groups.
The cars with a lot of buttons simply look outdated and people feel bad on choosing a car with a small screen.
The whole marketing is built on it, you get a small screen and lots of buttons if you get the basic version of the car and you get giant touchscreen if you buy the premium package.
If your new car has a large touchscreen your friends who own 5+ y.o. car compliment your choice and express jealousy(at the time of purchase, most people don't have real world experience with touch screens on cars and touch screens are in these cutting edge electronics that are expensive, so they must be good).
If your new car has a small screen you need to explain why this was the logical choice and how much you saved.
It's even the same with the iPhone 13 mini. That device is amazing, you can use it with one hand and fits in every pocket and the screen is actually larger than the first large screen iPhone(the iPhone 6) but people will try to understand why you bought that one. Are you poor? Why would you buy a tiny phone?
It's very strange, the word on the street is that the larger the screen the better. If your $30K product instantly becomes much easier to sell when you replace buttons with touchscreens without increasing the costs wouldn't you do that? I guess you need to have a niche, snobby traditionalist brand to be able to reject that demand from the consumers.
If your new car has a large touchscreen your friends who own 5+ y.o. car compliment your choice and express jealousy
Does this ever happen? I've never heard anyone express jealousy regarding not having a big enough touchscreen in their car. I've heard several owners of modern cars with touchscreens bemoan how complicated and slow to use they are. In my experience literally no-one who actually buys and drives cars thinks they are a good idea and many people - including myself - are deterred from buying a new model specifically because of the technology.
The thing is, it's actually really hard to judge quality of a design(takes a bachelor degree in Industrial Design and masters in related field and a few studies like the one in question to objectively evaluate a design). Most people like the new trendy one and unfortunately in cars that's a large touchscreen.
Don't think of car enthusiast, think people who like the car because of the shade of its color and feel of the leather - which is most people.
Don't think of car enthusiast, think people who like the car because of the shade of its color and feel of the leather - which is most people.
Those are exactly the people I'm talking about though. I'm in the UK - maybe the current culture is different here to some other places?
Of course it's also possible that my own experience hasn't been representative but I've heard the same story so many times for so long now that it's hard to believe I've encountered some freak sample of outliers.
I don't know for sure but on most brands you literally have to pay more to get the large touchscreen. Don't you think that the car manufacturers would put desirable features to convince the customer for an upsell?
Put in cheap features and then try to make them desirable definitely seems to be closer to the reality. Bonus sized cheap features definitely goes into that.
Also, chalk me up as someone who has never heard a positive thing about car touch screens after a week or so of interaction.
>The thing is, it's actually really hard to judge quality of a design(takes a bachelor degree in Industrial Design and masters in related field and a few studies like the one in question to objectively evaluate a design).
Horse hockey. Spend a couple years in Quality Assurance with your eyes open. It:s trivial to seperate wheat from chaff. The key that your Industrial Design might give you insight on is the fact that Industry has decided unilaterally that cost to produce > joy of end user in use. I.e. if it's cheaper to make and sell, it's higher Quality, rather than it's damn good, now lets streamline it.
Yes, your process weighs into it, but I assure you, the cognitive load of a haptic interface vs a touchscreen is so much lower it's absurd to even try to compare. If you really care about the end user, you take the time to get them buttons, and don't distract them with touchscreen finicky BS.
> Don't think of car enthusiast, think people who like the car because of the shade of its color and feel of the leather - which is most people.
Precisely. Millions of people (intelligent, rational, highly-educated) still buy cars with specific color/trim as their primary motivator. Until the trend reverses, a screen will continue to be a value-add to any vehicle because of the "modern" association.
I started buying base versions of cars to get away from the trend of shoddy touchscreens with bad software. Even if everything is done perfectly I get lost in them...which is not a safe feeling.
Honda has features they implemented in some attempt to streamline the experience but you still get lost easily.
After having a few vehicles with large touchscreens and then buying an F-150 XL to simplify, I can't even describe the elated feeling of operating a vehicle where the screen does what it's supposed to do with the vital controls all being physical. Yeah, I look like peasant but I get to keep my sanity.
On the contrary, rich narcissistic people already have the touchscreens in their cars. Those who wish their cars had a large touchscreen are people who can buy a new car every 5 to 10 years and they bough 2-3 years ago and didn't pay for a touchscreen upgrade.
I have a touchscreen on my 10 year old minivan. 10 years old and mini-van both loudly scream that this is not a vehicle that you buy to show off. 15 years ago a touchscreen was a novelty to show off, but now everyone has them.
Heh funny you mentioned the iPhone 13 mini. I just got a new phone and picked the iPhone mini. It’s by far my favorite phone since the iPhone 5. It’s also one of the cheapest new iPhones you can get. Like buttons on a dashboard the iPhone 13 mini is far and away a better product (for me).
And sells poorly, you can have hard time finding accessories for it because it sells poorly. Unfortunately, according to the leaks so far, it appears that there won't be iPhone 14 mini.
Relative to other iPhones, yes. I read that it accounts for 3% of iphone 13 (Pro, Pro Max, mini, standard) sales. The 13 line itself accounts for about 75% of sales. If Apple sold 40M phones per quarter, that 120M of the 13 line, so 3.6M of the 13 Mini. At 699, that's a 2.5BN business. Not too shabby.
I wonder if there's a way to make physical controls feel more premium via materials / design. In other consumer goods, there's definitely a market for physical design that feels more well-engineered with things like using metal and thoughtful trim. It's not surprising that people find black plastic buttons not particularly premium looking.
Definitely, I believe BMW for example had a big knob that just feels padded and luxurious. Same with car interiors; thicker padding, better noise insulation (e.g. when closing a door) makes things feel more premium.
It's somewhat interesting because I'm finding that I don't find touchscreens particularly "premium" looking. Touch screens and LCD screens seem to be everywhere nowadays in low-class places and look like obvious cost-cutting like Walmart and fast food drive-thru.
Back in the day, you could easily tell the difference between an expensive high-quality amplifier and a molded piece-of-plastic mass-produced boombox.
I think the way is to make them configurable. When you first setup the car you decide which controls will be "exposed" to the hardware knobs. In the winter time you may configure a knob to give you heated seats, during the summer you reconfigure it to provide max AC in one touch.
Car companies are already kind of doing this, usually just a button or two on the steering wheel. But IMO the entire dash should be a bunch of blank configurable buttons.
The complexity of modern car interfaces--including those that use random buttons--is already a massive pain in rentals. At least for nav, things like CarPlay standardize to some degree. But I frequently find myself hunting for all sorts of things on a rental.
BMW did this in the 7 series and I think some 5 series. A row of buttons in the center console that you can choose what they do. Unlike old style "preset" buttons they have a sensor in them to detect your finger being on the button before pressing it, and then it shows at the top of the iDrive screen what that button is programmed to do. I thought that was quite an elegant way to fix the "can't remember what I made this one do" issue you get with programmable buttons.
You really don't, and the fact that people consider it a given that you do says some very bad things about society. You should be buying the things that work the best for _you_, not the ones that will impress your friends.
> people will try to understand why you bought that one. Are you poor? Why would you buy a tiny phone?
I mean, fsck them. If I had people in my life who though like that (I don't), I'd get rid of them. If they're family and cannot be simply cut off, I'd minimise the contact.
The only thing sillier than judging people based on the size of their phone is cutting them off or "minimizing contact" rather than just explaining to them why they're wrong and moving on.
Sure, my laptop looks outdated with a keyboard. But compared to touch screen I'm more productive, faster, make less mistake, can wear gloves, don't have to look, can use it in sunlight, and it was never unresponsive.
Most people are not rich and car companies want to sell as many cars as possible, which means they need to sell it to the most people who are not rich.
5. Backup cameras. They're legally required in some jurisdictions (so I heard), and genuinely contribute to safety, but they require a screen. Once the screen is there, there's both less space for buttons and a virtual hook for features.
You can just use the screen for infotainment and satnav though.
I already have a screen showing what radio station is playing, and one on the dash telling me where to go. If a screen is required for a backup camera, just combine it. If I'm reversing I probably don't need the satnav anyway, whereas if I'm reversing or using satnav I probably do need other functions which just means you need an even bigger screen so you can fit everything on.
Sure, but the screen need not have any touch capabilities. It doesn't need to be large either. A 2 inch screen is large enough for the camera functions.
I have been surprised to learn that regulations only require that the backup camera turn on when you go into reverse; not that it stay on while you are in reverse. Some cars let you navigate away from the backup view even while moving backward.
There is so much opportunity for better regulation, without making more numerous regulations.
All cool until there's a wall or something like that starting right at the height of the roof. Can't see it in the camera but can in the mirror. Most cars with rear view cameras even warn: don't reverse by relying on camera only.
These days the instrument cluster is being displayed with an lcd. It could be used for the backup camera and wouldn’t be practical to be a touchscreen.
You shouldn't use your smartphone while driving because you might cause an accident while you are looking away from the road and not using both hands to hold the sterring wheel.
To be fair, while I agree that touchscreens are far worse for distraction than physical controls, they're far, far better than a smartphone. Phones are designed to hold your attention, have small text sizes and interface elements, require actually holding the phone vs just using the touchscreen, and a lot of distracted driving comes from wildly inappropriate activities like texting vs advancing to the next song on Spotify or something.
And the reality is that fiddling with the radio, fiddling with climate controls, looking at physical maps and written directions, etc. were all things long before touchscreens. (To say nothing of mobile phones, including before they got "smart.) Let's not pretend that distractions weren't a thing before touchscreens in cars came along.
> Touch screens are just their 60+ year old fossils deciding "that's hip, that's what kids want!"
Do you have any evidence for this? Seems pretty outlandish to me.
> 1. It's cheaper to produce; 2. It looks more expensive, so the price goes up; 3. Testing audiences respond positively to shiny lights; 4. Fossils decided that this is what the young people want.
What does 4 even mean? If we took 1-3 as fact, then should businesses have disregarded them and instead made something more expensive to produce that looked cheaper and sold for less because people don't respond so positively?
> What was even more surprising, to me, is that Mercedes had this amazing nice center console unit to control things with your arm in a rested position. They removed that piece of brilliance!
Mazda also has this beautiful dial-joystick which we can operate in a rested position. It is so intuitive that I stopped using the touchscreen console itself.
On the other hand even when we operate using a dial and buttons we take our eyes for an instant to look at the screen to check the changes. Now imagine looking away at a touchscreen just to see what operation to perform etc. This is a major distraction.
BMW at one point had (and may still have) a dial joystick, but I found it really unintuitive. Maybe it was poor software design, but it was never clear to me when I needed to turn the dial vs move the stick to navigate menus. Did you find it easier to use the control in the Mazda? Was the UX better?
Mazda has actually been going in the opposite direction by removing their touchscreens; 2016 Mazdas came with touchscreens, but the newest models go without. Personally, owning a 2016 Mazda I never actually used the touch screen once, due to as you note, the great dial interface.
I have a 2021 BMW and I really like the alternatives on button and touch screen it offers.
I barely use the touchscreen and when not needed I outright turn it off. This is quite easy because BMW has 8 buttons that can be mapped to any function in the touchscreen including turning off the main screen.
Another thing I enjoy is the gesture detector. It sometimes has false positives when I gesticulate a lot but it works when I actually intend it to. It is very satisfying to mute the radio or change an annoying music with a hand gesture. If they would keep trying to integrate and perfect it I think it would be the right direction for innovation.
Touchscreens are fine when parked or for the passenger. Anything else they are useless and often have too much distracting info, so they are turned off.
I ordered a new 3 series (2023) which does away with those buttons and a few others. I would have preferred having at least the temperature controls as physical buttons but other than that they do have the navigation knob/joystick which is well positioned to control the system. It feels like a reasonable compromise.
The joystick in BMW's is awesome. Gives you access to every function in the car with minimal glance at the screen needed. If you get the one with the HUD, no need to even look at the map.
Mine displays navigation and media information along with speedometer and street signs. It does not have all the car's info available.
Was looking at [1] and even with the joystick it is less than ideal. I would need to try though. The joystick is good to navigate the screen but navigating a screen while driving is still not the way it should be.
I can see voice control being useful and far cheaper (modulo the issue of how to deal with jackass passengers, but that's pretty easily dealt with using a push-to-talk button on the steering wheel).
In a previous discussion someone mentioned that part of this trend toward screens in cars is that new cars are now required to have rear-view camera. So once you are required to have the screen, it's really almost nothing to waltz over to touch screens. Of note: the "winning" car is so old it doesn't have rear-view cameras.
> What was even more surprising, to me, is that Mercedes had this amazing nice center console unit to control things with your arm in a rested position. They removed that piece of brilliance!
My MB is a UX disaster.
Have a guess how many controls there are in the car for navigating the (non-touch!) screen?
1? Nope. 2? Nope. 3? Yes 3. A touch surface in the centre console, a spinning wheel in the centre console (which is also a joystick), and finally a little joystick thing on the steering wheel.
Volume controllers? 2.
And don't get me started on how dangerously absurd it is trying to switch between MB's own system and Apple Carplay/Google Auto whilst driving.
> Touch screens are just their 60+ year old fossils deciding "that's hip, that's what kids want!"
Is there evidence it was 60+ year-olds who decided to lean on touchscreens for cars?
If that's just an assumption, isn't an equally likely ageist guess that it was pushed by people who came up through the ranks in the era of "UX"? (Since I'd expect that old-school, pre-UX human factors engineers, who grew up on research coming from aircraft cockpit optimization, safety, and UI in service of the user... would research the heck out of a new technology option like this.)
The parent comment could not have gotten it more wrong. It was not "60+ year old fossils" that made this decision. It was "30-40 year old disrupter hipsters" that told the older people in charge what looks immediately appealing to the average person (and not just young people, who can't afford to buy new cars).
My brother works in automotive engineering, it isn't 60+ driving this trend. It is the design team, which skews young, and the marketing team, which also skews young.
Tesla does not skew 60+ anywhere in the company, and they introduced these oversized screen based displays years ago.
So on you four bullets above:
1) True
2) I don't know, perhaps?
3) Maybe a quick 'image' audience, but are they doing usability testing?
4) Completely false.
The big weight is on point #1, for two reasons.
1) Those displays may seem expensive, until you actually price out the panels they are using. Then go and see what those physical buttons cost. They are not cheap. And there are a lot of them. And both technologies have micro processors behind them, so using physical knobs and buttons doesn't save money there.
2) Using modal displays to cover multiple controls saves dashboard real estate, and eases design constraints. Designers love it.
One of the things I hate the most, is that I want a mostly dark interior when I drive at night, and now I'll be stuck staring at an illuminated display that I hate using in any case.
More likely they are doing testing but aren't measuring the right things or are performing the tests improperly. I can say with high confidence that any of today's UX folks don't understand the scientific method nor statistics.
Or they know what they're doing, hate it but decide for it anyway due to some sort of FOMO (the competition does it also!) Maybe it's comparable to the glossy laptop screen fad some years ago.
* The one and only physical button car took 10 seconds total to complete their tasks
* Two touch-screen cars (Volvo C40 and Dacia Sandero) took only 13 seconds to complete the tasks
* Most touch-screen cars take 20-40 seconds
These results are certainly consistent with the hypothesis, "A moderately well-designed physical interface is likely to be better than an extremely well-designed touch-screen interface". But it's not really enough data to support the hypothesis that all physical interfaces are better than all touch-screen interfaces. You'd want to see what the curve looks like -- with it so close, it's quite possible that some, or even many, physical interfaces would take longer than 13 seconds for their benchmark.
And if you slow people down by 30% but reduce costs by a significant fraction, I think that's probably worth it.
I can guarantee you that they are doing exhaustive usability testing. I've had friends that worked in Ford's design and usability group. EVERYTHING is extensively demo'ed and discussed to death. My friends in the design group complained that the actual engineers would take their designs and fight them constantly on every change and that what WAS a nice interface was junk by the time it went into the vehicle.
I suspect that the engineers fighting them is really just a case of the hardware team and the software team not understanding the world the other lives in. The hardware team is working with a slow as molasses processor that is the only thing thats been approved for the ridiculously rugged life that a car CPU lives and the software people don't understand that just because a webkit rendering engine is completely fluid on their 6 month old Precision workstation it won't be on a 500mhz in dash processor.
Night driving is especially annoying if there is a lot of backlight bleed through the display. Perhaps OLED displays would make this better, but of course... more expensive.
It all makes sense from just the financial point of view. So that means it isn't going away any time soon, unless there's a huge backlash from consumers.
Perhaps the best thing we can hope for is 1 car manufacturer deciding: "Buttons first, touch screen(s) second."
Let consumers decide with their wallets. Though, I wouldn't be surprised that many consumers go for an inferior product just because it looks cool. Because that, unfortunately, is how humans work.
With touch screens, even the most simple task requires you to take your eyes off the road in front of you; with regular buttons you could do many task just with touch.
Sort of related, I have the exact same issue with portable music players while walking or cycling. Most of the time the only task I need to do is play/pause or forward/backward track.
For a player with buttons it takes a small amount of attempts and after that you've learned the position of the buttons by heart and can control the device even while it's in your pocket, without needing to see it. Usually aided by some tactile feedback. Fast, convenient, and somewhat safer since we're talking traffic situations.
With a touchscreen-only player that is much harder, sometimes impossible (depending on which screen you're in the controls might not be in the same place or not be there at all).
Sad thing is, this was already the case like a decade ago, leaving me wondering if designers have any pride in their UX, simply don't know they're doing it wrong, willingly just focus on other things apart from usability, etc. In any case: driving a heavy vehicle at high speeds should be the last case where simple things like switching a radio station actually requires you to take your eyes of the road. That's just insane.
Maybe I'm weird. I don't use the center console that much. For music, I'm either streaming or shuffling what's on my phone. And then I have the screen showing maps. If I want to put in a destination, it's usually done before I even get out of park.
My steering wheel has volume controls and the environmental controls are still button based.
You are missing the actual real value. It's flexibility. A modern car is software and gets updates. With a touchscreen you are least constraint by your previous assumptions and can change direction any way you like.
> You are missing the actual real value. It's flexibility. A modern car is software and gets updates.
The updates are only for bug fixing.Maybe this will change with SaaS but today i never heard of any Car company which does this. And no, i don't consider Tesla a car company.
I think points 1-3 are valid and certainly contribute to the decision, but many manufacturers believe that the future of the interface is a mixture of voice control (environmental, cabin lighting, navigation etc) and manipulation of steering wheel controls with HUD feedback (infotainment and everything else). Failure to embrace voice interfaces and demanding a button for everything is making 'fossils' out of 20-100 year olds.
Source: I work at one of the big German car manufacturers and have mostly drank the 'use voice, don't look off the road' koolaid.
> Touch screens are just their 60+ year old fossils
Please don’t play this game where a bad design decision was finally recognized as bad design and a scapegoat is found rather than admitting the “experts” who did it have no clothes.
UX branding itself “UX” rather than any of the half dozen other names we used to use was a clear statement of “It will be different this time, I promise.” It wasn’t. The design trends we got were different, but bad interaction design is still bad interaction design.
I'm not sure how you juggle the cognitive dissonance of cramming "Testing audiences respond positively" and "Fossils decided that this is what people want" into the same explanation.
I do agree with the premise that physical buttons and knobs are generally far superior to touchscreen UI's, at least for the common core basic things.
However, I don't agree that it's about the boomers, or the capitalists, or any other Internet strawman forcing something onto the masses against its wishes. I think it REALLY IS a matter of test audiences and "casuals" having tastes that differ from power users and other people that think deeply about a thing. You see this in many different domains.
I’m driving a rental MG on vacation right now and today I managed to hit a button that made the speedometer on the digital gauge screen go away. It’s just… gone.
Hoping it comes back after being turned off overnight, or I’ll have to find a manual online somewhere.
I know of at least two recent BMWs that were sold without handbook and instead it needs to be read on the touch screen in the car. It's terrible. Welcome to the future.
I mean, there's some positives to the digital manual to that assuming its done well. The manual will never get lost, it won't get wet, it won't get dirty. Searching on it could potentially be better than flipping to the index. Its not like it being on the screen is less safe or something, you shouldn't be reading the manual while you drive the car.
I don't agree with not giving you a paper copy of the manual though, especially when its a luxury car. The digital manual should be in addition to the paper one. They shouldn't be pinching pennies like that.
Oh yes, I wholeheartedly agree with you. There is nothing wrong with having the manual available from the entertainment system. But not having a paper copy at all is terrible. If I buy a new car I want to know all about it. I'd read that paper copy front to back straight away. I wouldn't do that sitting in the car scrolling through a screen though.
The whole thing with touchscreens in cars reminds me of the touch bar in MacBooks.
It's a nice looking gimmick if all you want is to show off but not suited for serious use.
One of the features of modern cars is conversational UIs (CUIs) that enable a truly hands-off approach from controls altogether, both touchscreen and analog.
I hardly use my car's touchscreen except for discoverability, and definitely not while driving. Speaking to the car is easier and more efficient:
- "I'm cold"
- "Turn on the windshield wipers"
- "Navigate to work"
CUIs have evolved to the point where controls are generally a hindrance to consumers and manufacturers both. Why include expensive, breakable analog controls when you can give the driver a better user experience hands off.
I agree with this. Voice control fitted to cars is pretty terrible still. It interrupts playback, takes a long time to resolve an input and takes even longer to repeat it back to you before doing it. Siri makes this a little bit better, but she's not been given access to the AC APIs of my car.
Buttons are the way. Or those cool switches from fighter jets... I'd like to see those in a car.
> Voice control fitted to cars is pretty terrible still. It interrupts playback, takes a long time to resolve an input and takes even longer to repeat it back to you before doing it.
I can't speak to all cars, but this isn't the case for Tesla. It's a very good user experience. It may be inconsistent across manufacturers right now, but as that evens out I don't see a barrier to more adoption.
Alternatives are important for accessibility, but I don't agree that they need to be available to all drivers at all times. The idea of going back to panels of analog controls that I have to reach for and memorize the positions of just isn't appealing. They break, the paint fades, they chip. CUIs are easier and safer in general.
Start with telling it what you want in natural language. Conversational AI is good at understanding variations like "I want, " "I need," "please," etc. The touchscreen can help you discover what features the car supports, and those should have CUI commands that are easy to intuit, which you can then use while driving.
Sorry, but this sounds like a nightmare to me even if it worked well, but the actual quality and robustness of CUIs is nowhere near what you imply. In any case, there’s about 0% chance any car computer is going to do the right thing if I say “I’m cold”. And in no world is saying “turn on the windshield wipers” a better user experience than turning a knob within easy reach, especially if you are listening to anything or if you have other people in the car.
> the actual quality and robustness of CUIs is nowhere near what you imply
I don't agree - I've been very happy with my car's CUI. I can't see going back to analog controls or touchscreen. There just aren't that many commands that I need to execute while driving, and the CUI understands them all.
Clicking a button is orders of magnitude faster than casting a spell. Voice controls fail more often than analog controls break. You do not want a delay to turn on your wipers when you are at 60mph and hit a sudden localized rainstorm.
> Clicking a button is orders of magnitude faster than casting a spell.
It depends on how many buttons there are and how familiar the driver is with them. I like not having to remember/look for the positions of analog buttons on the dash.
> Voice controls fail more often than analog controls break.
I haven't had a voice control failure, but any system can be designed with varying degrees of reliability. CUIs have a lot to offer in the way of safety and convenience.
> It depends on how many buttons there are and how familiar the driver is with them. I like not having to remember/look for the positions of analog buttons on the dash.
If it's your car, this problem will solve itself for you in short order. There's not that many functions that a dashboard needs to do that you won't familiarize yourself with it in a month or two of driving.
Look to a computer for an example, some shortcuts that you probably use on your keyboard are downright arcane, but because you use it so frequently it's probably natural.
My car has full touchscreens and can't understand even the most basic voice commands. Ask it to navigate to the town of Drachten (NL) and it's likely to route you to some village in Serbia or just say "i don't understand".
It's utterly useless.
"turn of windshield wipers"
What speed? I know intermittent. I don't know what the fast speed is called, I call it crazy wipers, I suspect the manufacturer's don't call it that. And that's just a 3 speed wiper, some have more speeds.
"I'm cold" well if you say 'goodnight' to Alexa, she says goodnight and carries on with the music which isn't what I want. I can imagine "I'm cold" being the same, and even if the car does get the hint what do they set the temp to?
"Navigate to work" I know where my work is, I mostly want to navigate to places I don't know, and are therefore not in the memory. So it's "car navigate to some street, city X"
"Would you like 1 some street, 2 some street, 2a some street...."
Or
"Do you mean some street or sum street or summ street?"
And that's assuming they know the pronunciation, the locals of Slaithwaite can't agree on the pronunciation. And theres many places pronounced weirdly.
Mine triggers when I say my kid's name with different vowels and number of syllables. I had to turn it off.
If I have to press a button to make it work I might as well press the button that does what I want.
It seems like a good idea for adjusting the climate control but I apparently didn't pay enough and it replies something along the lines of "I'm sorry Dave, I'm afraid I can't do that."
So far it hasn't offered a monthly subscription to enable that feature.
I'm not sure they would work well if you're listening to loud music or something. Personally, I don't think I'd ever buy a car that uses voice control as its primary interface; I greatly prefer physical buttons.
It actually does work on my Polestar even when music is played, it pause the music while listening. Polestar uses Google voice recognition which is great, but still I hate using it except for rare cases at it is slow and prone to misunderstandings. It is ok-ish for listening to incoming messages though.
Good lord, no. Voice control is the worst possible UI, because it’s just as easy to flub or mumble or be misunderstood as it is to aim wrong on a touchscreen. It’s far more vague than actual buttons, knobs, and touch screen controls. How to even describe what you want is a challenge, or even knowing what’s possible. Plus if you are listening to anything or have other people in the car, it’s incredibly disruptive, not to mention the risk the car will misinterpret something actually being said.
Agreed, I use scroll wheels on my steer for volume and next/previous song. Voice control for temperature & navigation and if an text comes in do a quick reply.
Honestly I barely feel the need to use the touch screen in my car while driving, and if so I always make sure autopilot/high way assistant is turned on as extra safety.
Familiarity is my personal favourite part of driving. Knowing the road, how much input you need to apply to a turn, knowing where the buttons for things are, just being able to feel for a control and know its purpose. All while my eyes are on the road.
Same situation with even modern land cruisers, with the exception of the hvac system in the top end models. A common modification is to order the base model modules from Australia and get rid of all the touch screen components.
And the new generation are all using touchscreens which need you you take eyes off the road to find the fan settings (return to home screen, touch near the bottom centre, then find the wedge-shaped fan speed widget and adjust that). Temperature needs to pop up another pane and set that by clicking the temperature number and then manipulating a bar chart thing. None of this except the "return to home" button is tactile in any way.
Before: turn the dial. Done. Temperature is the one next to it.
Adjusting the sound balance is downright dangerous (pull down the Android-esque menu and click though levels in the UI). Old car: press the tactile centre of the volume wheel until it says balance, then use the wheel.
I think this is more a problem of standarisation and if car companies got together and pitched in a bit of research to find out what worked well with users they could all get an interface (or a set of standards) that worked well for most users.
If we had to dial every number we want to call manually, physical buttons on phones would outperform touchscreens too. But nobody does that anymore. And it is nicer to select the name of a friend on a touchscreen than to type it with physical keys.
Same goes for cars. Humans spending their time keeping the car in lane, stopping at red lights, starting when the light turns green, doing turns etc will phase out more and more over the next 10, 20 years.
The interaction we still want to have with a car is probably nicer on a touchscreen. Especially when you are not dabbling with a steering wheel anymore.
Stuff like seeing the route on a map, selecting waypoints etc. We would not want physical keys for that when we are at home at our computer, right? I think dedicated keys are just legacy from the "I'm busy with the steering wheel and need to do other stuff blindly" area.
Much of it will probably also move to voice control.
Any quantifiable arguments that 10-20 years is aggressive?
Google started development 13 years ago and now already has fully autonomous taxis on the road in a 4 cities. If they double the number of cities every year, in 10 years they will be in over 4000 cities.
Tesla started around the same time. Looking at videos of tesla in self driving mode, my feeling is that without human interaction, it would crash maybe once every 50 hours of driving. Double that every year and in 10 years it will crash once every 50000 hours. A human driver, driving 1 hour a day, would need 136 years of crash free driving to achieve that. But the average driver has 4 accidents in their lifetime. So we are already in superhuman territory by then.
I guess my argument against the aggressive timeline is that I have been working in MV/AI for the last decade or so, and am pretty deeply connected into NVIDIA, AMBA, a few SS lidar companies and so forth (though I do not deal with autonomous driving, I am in physical security/surveillance).
While there has been a lot of progress made to date, much of it (IMO) has been around the low-hanging fruit kind of stuff. Sure, we've mapped a lot of roads and covered many of the basics. But none of the systems are really able to do human-level context awareness, like you see a bunch of kids running around near, but not on, the road a few hundred feet up, might want to slow down or at least be extra aware, or people clearly driving in "tourist mode", etc.
My personal assessment is that we are still 20+ years away from the point where the human driver is mostly along for the ride (and thus to the point of this thread, can fiddle with a poorly implemented touchscreen UI without risk). I am led to believe the ultimate solution is going to involve altering/enhancing the road infrastructure, along with the general improvements for the in-vehicle stuff (which itself has at least a decade of development still to go). Combine those things together, and we're still quite a ways off from the vision that was sold 10 years ago.
So the "argument" you give is your gut feeling when you look at how things are now.
I expected something like this. That's why I said "quantifiable". We humans are animals of habit and have a hard time imagining a changing world. But every time a few decades pass, we look back and see - damn! - a lot has changed.
A way to get away from feeling and towards a rational prediction is to look at rate of change. In my experience, doubling the performance of a new technology (self-driving is only 13 years old) is usually doable. And that means 1000x improvement in 10 years. And from where we are now, that gives us Waymo in every major city and Tesla with super human capabilities.
So your response is essentially your gut feelings then?
I do not think your rate of change assessment is accurate. With many newer technologies, like the DNN/CNN frameworks, we see a more or less immediate order of magnitude improvement, and then declining rate of improvements as the easier bits are done and we are forced to do deeper development or refinement.
Saying that Tesla has super human capabilities sounds to me like you are not assessing it rationally. Super human would mean, to me, that it manages to do better than humans in extreme conditions. Instead, we have seen a number of very preventable causalities in the current generation systems.
What evidence have you that growth in self-driving capabilities is inherently exponential, much less doubling every year? Indeed, if I extrapolate your numbers into the past, when Tesla started self-driving it would have been crashing minute or so, which is laughable.
The amount of data grows exponential. Not only do the existing Teslas keep adding data, but more and more Teslas are added to the fleet, accelerating the pace at which data is generated.
Crunching the numbers gets exponentially faster, because compute power is growing exponentially.
The algorithms used to crunch the data become better.
As more and more revenue comes in, more and more can be spent on data crunching.
The sensors become better.
Maps become better.
All of these factors multiply. Adding another exponential force. Even if the data would grow at a linear pace and the algorithms would get better at a linear pace, this would result in exponential improvements, as these two factors multiply.
The click wheel is a very good interface, especially when you are very often setting a scalar value (volume, temperature, fan speed, etc) or navigating a linear list (songs, contact names, call history, etc) and using one hand without looking at that hand. So it's a good fit for both MP3 players and cars.
Not so good for typing, but can be quite good when there's a good UI that narrows down options for you at each entered letter.
Voice control is not an improvement imo. An anecdote...
I wanted to make a call to my partner recently, whilst driving. I have an old car, with all physical controls. But my phone was in front of me with maps running. I said "OK Google.." and waited for it chime that it registered. "Call [my partner's name]", I commanded it. It responded that the name was unrecognised - even though it repeated the name back to me, so it wasn't a case of mishearing. I repeated this at least 5 times, getting increasingly frustrated.
Then eventually it started calling. Joy! But I notice it doesn't say my partners name on the screen, just a phone number. "OK", I think to myself, "perhaps this is some quirk of voice dialing, it doesn't show the name" (I don't know my partners number by heart so I could not recognise it).
It rang through to voicemail, and whilst the name given by the voicemail recording was correct it was definitely not my partner's voice. It turns out my phone had just dialled a random mobile number from the Internet that matched that name. It may have been a business, or a personal number, I'm not sure. Either way, I would never want this behavior!
I guess the point is that voice commands have terrible error handling and recovery modes. Give me physical controls any day. Even if they're slower, at least they are accurate, discoverable, and do not make guesses about intention
Perhaps when driving is no longer the main focus of the driver in the car, a touchscreen interface might make sense. Except that is not the case for any car on sale right now in any state of being (save for being parked and motionless). Yet there is still much eagerness to move everything to touchscreens right now...
981 comments
[ 2.7 ms ] story [ 361 ms ] threadBut after a while living with the stupid UX, the buyers would probably rather have the functional than the shiny.
But cars without roofs are getting out of fashion anyway. Next step after no steering-wheel will be no windows.
More generally: I want to be able to feel the state something is in. Whether it is push/pull, turn, or flip doesn't matter, but it has to physically maintain the state I put it in.
An exception to this of course are functions that are used temporarily, like the horn, and the wipers swipe-once. And if direction indicators turn off automatically, the input device has to change its state accordingly.
I’m a flight sim nerd and it’s immediately clear when you have to keep track of more states than your are used to how much attention is needed and how much is freed up if you can reach over and feel the state of the switch.
I would love that for many functions in a car. And a knob with some bump or whatever at the ac temp setting, so I can adjust its position based on feel alone as well.
On a modern keyboard the caps-lock key would probably benefit from one too =)
I think switches that can set themselves to current state (flip back) are a bit pricey but can be had from $12 or so (iirc).
I hope the automobile manufacturers take a note of this, and bring back some of it that’s worked for decades. I believe few German manufacturers like the Mercedes-Benz to have a sense of this, hope they don’t get along with this trend too.
Joking about the price aside I always loved that the original design goals for the F1 included that if someone found one in a barn in 50 years time they should be able to repair it.
I think that leads to quality, simple, tactile controls.
The exception would be to put a tiny motor on each control to keep it synced up, but the cost to reliability (not to mention just money) would be unacceptably high.
Power windows are close to fulfilling this (up/down separated, and it might be argued that the state is sufficiently obvious, though questionable for the rear windows). A central lock controlled through a common lock/unlock button does not satisfy this.
I think the big sticking point is usually climate control and radio since that's what people fiddle with most while driving. Maybe little things like the ride mode selector (but I'm not sure how many people actually fiddle with those sorts of things in practice). But nowadays the "radio" is rarely just radio or disc/cassette player. It's music, audiobooks, or podcasts off a phone. I don't really know how you do that with integrated controls that aren't screen based, like most luxury cars already have the little control wheel and also controls on the steering wheel. So what more would you need? Maybe some haptics on the control wheel to give you more feedback?
Honestly it makes driving a real chore and is a huge safety problem; just that one little anti-feature ruins driving.
Most cars, most of the time, are driven by people who are very used to their particular interface. I couldn't find mention of this in the article, but it's necessary to state.
Furthermore, increased automation in many of these cars means you don't have to do many of the listed sequences as often, or at all:
"Lower the instrument lighting to the lowest level and turn off the center display." Why would I ever do such a thing when this is fully automated already?
> One important aspect of this test is that the drivers had time to get to know the cars and their infotainment systems before the test started.
Why was that switch there in the first place? Being familiar with the car didn't help her remember it existed, even when she encountered the "windows wont move" problem repeatedly. Assuming you want a "stop the windows working" switch, why put it up in the corner of the dashboard and in a form where it's easy to hit accidentally and hard to tell that it's not in the right position?
The problem is deeper than screens and older than "legally required backup cameras;" the ergonomics of the car interior became secondary to "but marketing says people want feature $X" checklists.
It gets others, too: https://www.insightcentral.net/threads/passenger-window-stuc...
Ahh, that is a high touch area. Thanks for the response, I was worried that my question was a bit blunt.
To verify that a mountain was indeed being made out of a molehill, I checked the manual [1] for the 2002 Insight: page 78 explains this is the case in a single sentence.
[1]: https://techinfo.honda.com/rjanisis/pubs/OM/AH/AIN0202OM/enu...
Tractors and boats have amazing layouts that accommodate your engagement in operating. Cars are full of gimmicks and showroom fluff
If you look at Gordon Murray's latest car, the t50, he spent a lot of time sourcing buttons and switches for the car because he cares deeply about their tactile feel. There is also not a single touch display in the car. Because he passionately hates them.
If you are into sports cars you kind of want to cry when you look at how badly some of the dashboards tend to age. Some of them look like embarrassing student projects where a 1990s web designer has crammed all manner of animated gifs with a horribly infantile palette onto the dashboard.
That's part of why I have an Audi.
It's a 2021 model with a touchscreen. But it also has high-quality buttons and knobs for each of the tests in the article mentioned and would thus pass with a very low distance measured.
There are some weird issues, however. The audio buttons control the current audio source but don't fall back if the current audio source disappears. Like if you are playing music through your passenger's phone via CarPlay. If you drop them off at their house and keep going, the physical audio controls do nothing at all until you select a new audio source on the touchscreen.
There's a reason professional cockpits still largely eschew touchscreens when 250+ lives are at stake in the back.
I bet Lockheed Martin sells a $50K glove that works with the F-35 touchscreens.
It's not that bad, however. In a car you need to be constantly aware of the road ahead of you. On a plane, you are not required to have your hands on the controls while on autopilot and you can pay attention to the screens, as well as operate them - there won't be any wildlife crossing ahead of you, not any red lights forcing you to brake.
And if the plane sees a threat it will warn you well before your human senses can so you can pay full attention to it (and the helmet-mounted displays). I assume the F-35 also has stick-mounted controls to operate in the helmet display.
I predict in 10 years car manufacturers will bring back physical controls for some things (if not all).
One advantage physical controls have is that I can operate them sightless once I learn the layout. Most touchscreens I've seen in cars don't really have that feature because of the design of the system behind it (whole screens shift so returning to navigation isn't often simple, for example).
Maybe. If new cars are all self driving, I could see manufacturers keeping with touchscreens. It's safer to stay distracted longer if you're not in direct control of the vehicle.
Eventually if we get to actual full-self-driving they might go that route. That's way more than 10 years out though, I think. You're talking about a point where cars are able to not have a driver at all before that becomes a reasonable option IMO. That said, even elevators have buttons.
Not hard to make an insertable switch component that snaps into a slot, and which can be removed and replaced in 10 seconds using a simple tool (or maybe no tool at all).
The flat screens look good, and work great in calm situations, but in heavy seas and/or rain, they can be challenging at times. In my case the manufacturer (Raymarine) offers a wired remote control, so I have a knobs-n-buttons controller that is easy to reach and offers more direct control when needed.
I've heard from friends in the aviation industry that pilots take extra care to put as little stress on switches and buttons as possible during normal use to prolong their service live.
The uninitiated might think why bother when a switch or button is dirt cheap, like several cents per unit cheap. And they would be right, the best kind of right. But when a switch/button does inevitably fail and needs to be replaced, the cost can easily come out to at least several hundred bucks between the labor, reinspections, and recertifications among other red tape that help ensure safety.
So if (if!) touchscreen interfaces are more durable and last longer, that is one fair argument in favor of them over mechanical switches and buttons.
Before a touch interface fails from long-term use? I highly doubt it. Plus, a switch is a very easy component to replace. Touch screens can be but aren't always.
I think service life arguments are just poor effort. We well understand the appropriate average service life of a variety of switches. We don't understand the same for touchscreens, especially modern ones, as they haven't been around as long.
From experience, in aviation nothing is simple or cheap to replace. Cheaper than replacing the entire Garmin unit though, yes.
Of course it is relative but I would still assume a switch in aviation is easier to replace than a touchscreen.
The main ones (autopilot and flight controls), rarely fail if at all. The only ones failing from time to time are small switches for radio channel volume or cockpit light adjustments and system buttons at the overhead panel that are easy to replace by maintenance.
Touchscreens are not a good option for main controls due to poor visibility(dirt from fingers and sun reflections), hidden submenus, turbulence making hard to press the correct button…
The A350 and 787 are using trackball controlls for submenus and the onboard computers, not a touchscreen.
This is one of the biggest reasons I dislike touch screens in cars, yeah. Tons of roads are more than turbulent enough to make it hard to hit buttons. Not having a physical edge / clicking / etc to tell you where you are and when you've done a thing means you have to use your eyes, which means disabling what is by far your biggest safety tool while driving.
I remember once driving, I think a Renault Megane, which had all the controls (stereo, climate, etc) replicated as buttons on (behind) the steering wheel. A slight learning curve, but completely seamless driver-car integration once learned. Someone had obviously thought carefully about how this should work, instead of just slapping in a touchscreen and a bunch of menus.
Alternatively, what do you mean, quit?
Other than that - it worked and somehow had all the needed papers to be street-legal.
The "crapification" you describe is the creep away from the scientific principles that once underpinned this field. Before UX we had HCI (Human Computer Interaction) which was in turn a development of CE (Cognitive Ergonomics) and other "human factors" sciences.
These sciences were rooted in very rigorous but time consuming tests, observation, psychology and physiology.
from TFA: "Designers want a "clean" interior with minimal switchgear"
This is where the wheels fall off the wagon. Should "what designers want" be high amongst the priorities for safety critical products?
[1] https://iea.cc/what-is-ergonomics/
> tests, observation, psychology and physiology
Is that not their job?
Buttons and actual things that let you use the device/vehicle for 8+ hours a day, not so selected for.
I've only tried it in a VM for a few minutes so far, but was unnerved by the general feeling of 'pretty, but impractical', mainly thanks to the taskbar and the right-click 'hide everything by default' context menu.
Evidently not if the work they're producing is reportedly outperformed by old school physical controls from more than a decade ago and in most of the vehicles tested it wasn't even close.
UX usually focuses on the critical path for the top-5 tasks. So turning on the car radio makes sense, but changing the radio station didn’t make the cut, so some rando engineer guy stuffed it in a menu.
When it’s done well with a great team and time it’s magic. It’s easiest to see when Apple gets software right, like Keynote - the functions of making a presentation are immediately obvious to an elementary school student. But even then, once you leave the happy path, woe to you - modifying a template is a dark art to most people.
I'm convinced most people really don't like iCloud, but since the alternative is iTunes, they basically have no choice...
iTunes is the equivalent of legacy VB apps in enterprises. As far as I can tell, there was essentially no design for many years.
Apart from the remit being just too broad, designers in any case are part of a complex team that deal with a multitude of functional, non-functional, regulatory and financial requirements.
Now, we have many different definitions of "designer", which I am very aware of, but I believe that, in some circles "designer" has become romanticised and extended to include a set of perceived "magical" powers to "deliver what a boss wants". That is a distortion of the role to something grotesque.
Speaking from a domain in which I have expertise; in sound design a great battle ensued between designers, users (audiences) and the 'bosses' (studios and publishers) as to how music and films should sound. You probably know this as the "Loudness Wars". I think it remains a textbook example of misalignment between technical, artistic and financial factors. It also remains an example of why I think "Markets are a myth" [2].
Despite listeners saying over and over that they "Don't want it", the producers, through a mess of internal motives (mainly financial), repeatedly foisted their values onto them, being obsessed with what they think users want in preference to flat-out contradiction that would be evident in even the most cursory market research.
The job of a designer is to balance factors, and in a sense act as an advocate (stand-in) for the user by mentalising their actual needs. It's a very demanding and complex skill. Doing "what your boss says" is absolutely not it and reduces a designer to a tool.
On the other hand, a job of the designer is also to listen to expert technical advice outside of their skill-set, and so must not get carried away with any grand "aesthetic vision", wanting to be Steve Jobs.
A hard line to tread, and one requiring strong will and ethics as well as judgement.
[1] https://en.wikipedia.org/wiki/Loudness_war
[2] https://news.ycombinator.com/context?id=32463541
https://www.slashfilm.com/673162/heres-why-movie-dialogue-ha...
Designers can do all those things, but often they're not given the space to.
The best products are typically produced in an environment where the people running the company care about the design. This is a rare environment.
Speaking as someone with an Interaction Design (IxD) degree: no we fucking don't. Tactile buttons being superior has been known for ages. For example, Bret Victor wrote "A Brief Rant On The Future Of Interaction Design" in 2011, so over a decade ago[0]. Not that anyone with the power to change things listened, because these decisions aren't made by the designers.
This is mostly a consequence of people higher up trying to save costs by using touchscreens, which is cheaper to buy and cheaper to develop for. HCI and IxD have always had this issue that we're asked to fix things up after everything else has already been decided. Basically, we're mistaken for graphic designers who decide on what the final product will look like. So we're given a touchscreen to develop an interface for, not a blank-slate car interior (or whatever) for which we get to decide the button layout.
At the risk of pulling a "no true Scotsman", this is a consequence of cost-cutting first and foremost. Don't blame the people who actually have a background HCI or Interaction Design. We all knew this was coming, and we hated it. If we're told to make do with the touchscreens we are given, with the alternatives of actual physical buttons being ignored before we even get to make decisions, then don't blame us for the lack of those buttons.
[0] http://worrydream.com/ABriefRantOnTheFutureOfInteractionDesi...
For example, we're not talking about one button, we're talking about a lot of buttons, usually custom-made for the car in question. The whole dashboard physically has to be designed around them. Meanwhile Tesla just slaps a screen on a mount in the middle of the car and calls it a day. It's basically "we have to get everything right the first time" vs "fuck it, we can always fix things in a later software update". Which is a way to save costs by cutting corners.
The buttons all have their own complicated logic too, although I suppose that even with physical buttons one can handle almost all of that purely through software these days, so that's not really as much of an issue any more as it used to be (it does make me terrified that cars can be hacked and bricked, but I digress).
Speaking of a lot of buttons, that's the other thing: if all your buttons are virtual, you can have infinite buttons! The only thing we have to do is introduce a ton of mode switches! Which is absolutely terrible when you're driving, but nobody seems to care! So we can cram a ton of features into a screen that would otherwise require a million buttons, and use that in marketing. Even though we'd probably be better off if some time was spent to whittle things down to the essentials and design the interface around those cleanly.
A touchscreen is largely just a single fused physical unit with ~two cables: a data ribbon and power. Utterly trivial to install and wire up in comparison.
The total assembly cost adds up very quickly.
I don't think we're going to get physical interfaces back until car manufacturers (or whatever) are forced to because of said cost-cutting.
Note that "customer" can be a manager or similar higher up in the hierarchy.
Some remarks stood out for me:
> talk about technology. That's the easy part, in a sense, because we control it. (my emphasis)
Yes, I agree with him strongly. But - there's been a dreadful anti-intellectual tide this past decade - a descent into "technological determinism", or the idea that technology is its own process to which humans must bend. It's the idea that we don't control it. It comes along with the overuse of words like "inevitable", "ubiquitous", "unavoidable" and endless talk of cats escaping from bags and genies refusing to go back into bottles. It's a defeatist and lazy creed that seeks to excuse a race to the bottom of cheapness, as you describe, with a narrative about how we "have no choice".
> if a tool isn't designed to be used by a person, it can't be a very good tool, right?
Increasingly, tools are designed to be used by other tools. Humans are being sidelined amidst the interplay between machines. For example; the demise of the Web is largely due to bots and the arms race to create other gatekeeping bots to defeat them.
> Hands
Bravo! Not "a finger" or "your thumbs". That's why I use a keyboard, interact through text-based technology, and cannot fathom thumb-twitching smartphone users. I totally get what he's saying, having worked in sonic interaction design with musical instruments (NIME) stuff like the ROLI seaboard (or whatever they changed the name to)... hands and touch, with mechanical haptic feedback is the way to go.
I wish more people payed attention to this understanding of our relation to technology as embodied beings, instead of chasing a "clean" disembodied dream - which I think hides within sublimated Orthodox Dualism in the tech community - but that's another story.
The soft top folded down pretty nicely. The rear seat could also fold down, as did the windshield. The half doors were still removable, and there were still only 2 of them.
It was a manual transmission but had A/C. The heater had two settings: "lava" and "off" :D
Very unrelated.
Now, you control your car, though with increasing friction.
Eventually the usb port just stopped working so I use VLC on the phone and stream to the car via bluetooth.
https://news.ycombinator.com/item?id=32162131
I just want it to recognize that I sort my music into folders and show the folder hierarchy. Instead, in my last car, it would flatten the entire structure and then sort by filename. To top off the shitshow, it would also only show up to 99 files, and scrolling through the list was painful. Each tap on the scroll button would only move one line, not one page, so if I wanted to play the 80th song, I'd have to tap to scroll 80 times.
seriously though, the USB audio player is a dumpster fire. They completely redid it with the last update -- and it got worse, not better.
some random examples off the top of my head: - switch to another input source and you lose your audiobook position - sorting is completely utterly broken - audiobook has no skip-back-a-little function - soooo many bugs
The only nice things I could say is that it mounts linux filesystems plus it plays FLAC and Apple lossless directly.
I think Apple's design choices provide a better analogy but the point is still taken and I tend to agree. They want cars to be disposable technology that consumers are continuously upgrading (like their phone)
That being said I don't actually use any features during driving, except for adjusting climate control and compulsively checking fuel economy.
I'm curious what features people miss now that physical controls are mostly gone.
Adjusting the radio station is the other big one.
Driving a car with touch screens (new BMW or Mercedes) has left me very unimpressed. My 2016 VW Golf has actual buttons, switches, and knobs to twist and turn and press and flip.
Car reviewers, too, often say it's a shame that car manufacturers are switching to touch screen nonsense. It's such a shameful trend if you think about it. The BMW series of pre-2022 had buttons in the dashboard, but the upcoming new series will do away with those entirely.
Touch screens even find their way onto steering wheels and doors.
Of course, it's easy to understand why:
1. It's cheaper to produce; 2. It looks more expensive, so the price goes up; 3. Testing audiences respond positively to shiny lights; 4. Fossils decided that this is what the young people want.
Honestly, I hope European legislation makes it illegal at some point. For the sake of safety. With touch screens, even the most simple task requires you to take your eyes off the road in front of you; with regular buttons you could do many task just with touch.
What was even more surprising, to me, is that Mercedes had this amazing nice center console unit to control things with your arm in a rested position. They removed that piece of brilliance!
So, now you need to do everything with an outstretched arm in a moving vehicle to operate tiny buttons on a flat touch screen.
Oh, and the touch screen can only barely hit 60 frames per second and often feels much slower. They're even saving costs on GPU power in their fancy luxury cars.
It's mostly a cost saving measure.
Physical buttons are expensive. I you eliminate them, the car gets cheaper to make.
That's all.
It's a sign of low quality and I expect that in 5 to 10 years, consumers will start to realise this.
There are also reliability expectations. If I need defrost because the windshield just fogged over, it better activate when I turn that knob on my sixteen year old car. And so far, it always has.
Years ago when I was working in IT at FoMoCo I recall seeing a piece of paper on an office bulletin board outlining how they had managed to save like $40 on the production cost of a Taurus, a vehicle that at the time was about $20,000. Those savings were the result of multiple sub-$1 to several dollar cost savings tweaks made between production years.
Ford has built something like 8 million Tauri, save a few dollars on each of them and it adds up to real money, like enough to redecorate the executive cafeteria.
Gluing a single tablet is much faster.
My favourite feature of touchscreens in cars is when you try to click a button but go over a bump[1] so your finger misses and you press something else. Genius.
I do get that touchscreens allow manufactures to add and remove controls though.
Rolling out UI changes for self-driving cars, like getting rid of the, knob behind the wheel, will help with safety no end.
1. Not sure what the bump was, probably the neighbour's kid or dog or something. Too busy trying to get the latest Smartless. That Will Arnet, what a card, etc, etc...
Consumers is the key word here. Manufacturers already know that buttons have the "disadvantage" that they break independently. They are also easier to fix with a generic replacement part. Which means that you won't have to scrap your car because it suddenly became unusable.
Force manufacturers to provide replacement parts for 25 years after original purchase, and see them flocking back to the basics. But that prob. won't happen in EU (because it's against the interests of Germany) or USA (because "communism"). So I guess we're depending on the common sense of Japanese and Korean manufacturers?
I'm not so hopeful. The same can be said about household appliances. Yet more and more random things figure they should have touchscreens, or at the very least touch buttons. And this trend has lasted for far more than 10 years.
https://www.carbuyer.co.uk/tips-and-advice/170098/bmw-idrive...
edit: Nevermind.
> The main part of the BMW iDrive system is a control wheel, which can turn clockwise and anticlockwise like a volume dial. It can also be pushed forwards, backwards and to each side as if it were a joystick, and the centre acts as a button that can be pressed to confirm a choice or select an option. As mentioned above, later versions have adopted touchscreen technology, gesture control and voice commands, so there are multiple ways to operate a newer iDrive system in addition to the rotary control.
It's a physical user interface with buttons and a joystick. Which is basically what everybody here wants.
If Toyota half-way through shipping their latest car realize it's better to have two knobs on the dashboard, they can very easily add one if the dashboard is just one big touch screen.
With touch screen and OTA updates, you can skip the hard part and leave it for future you to improve if needed. But as we all know, when it's already sold there is no motivation to spend money to improve. So touch UI stays half baked. And only gets improved with future models.
The irony is of course that the decision to have very few buttons (not one, not zero, but very few) with almost all input via the screen was made very carefully by Apple with very specific justification based on understanding of how phones were used and could be used. This is clear from Jobs' iphone keynote.
If Steve Jobs, Jony Ives etc were redesigning car interfaces it's far from obvious that they would have made similar decisions.
No doubt it'll get to the point where you can't update it any more - either due to hardware incompatibility, lack of processing power, or some new technology being added. But it has meant that a 2013 Model S looks more modern today than it would have otherwise.
Equally, tech tends to look dated much faster than physical buttons do. It's too early to really say which has more long-lasting appeal.
[1] https://youtu.be/TZ0HsN-tblo?t=124
"UI revision 1.23: move the Passenger Seat Blender switch further away from the air conditioning controls"
A car is a dependable tool. Changing the UI during a car's lifetime is dangerous and unprofessional. I'd say the same is true for smartphones and computers but I guess the majority of people think of them as simple "cool entertainment devices"
this is a unproductive conservative attitude
no I won’t get off your lawn
I’ve been burned by changing UIs but it’s the price we pay for progress
tell me more about how updating your car's UI overnight can make you more productive
all updates taking together will allow for progress
This is a bug, not a feature.
If I'm driving then I'm driving. I want any non-driving controls to be as simple, consistent and reliable as possible. I don't want any non-essential controls at all. I don't want anything I might want to use while driving that requires me to take my eyes off the road at all. I couldn't care less what some flashy touchscreen UI looks like because I should never have to look at it.
The physical controls on the dash of every vehicle I drive regularly still work as well and feel as comfortable to use as they ever did. In some cases those vehicles are over a decade old. I'll take that over the modern touchscreen junk any day.
Sure, on a minute-to-minute timescale, anyone must obviously agree.
But over the long term of owning a car, it is an immensely valuable feature. My 2018 car still feels quite new and fresh - much less reason to replace it than if it were falling behind.
There's a substantial group of people who lease cars, and just get a new car after the lease is up. For that group of people, that's probably a substantial reason along with the exterior styling.
The car won't update itself, so:
1. Ignore any updates as they become available.
2. Problem solved.
The whole marketing is built on it, you get a small screen and lots of buttons if you get the basic version of the car and you get giant touchscreen if you buy the premium package.
If your new car has a large touchscreen your friends who own 5+ y.o. car compliment your choice and express jealousy(at the time of purchase, most people don't have real world experience with touch screens on cars and touch screens are in these cutting edge electronics that are expensive, so they must be good). If your new car has a small screen you need to explain why this was the logical choice and how much you saved.
It's even the same with the iPhone 13 mini. That device is amazing, you can use it with one hand and fits in every pocket and the screen is actually larger than the first large screen iPhone(the iPhone 6) but people will try to understand why you bought that one. Are you poor? Why would you buy a tiny phone?
It's very strange, the word on the street is that the larger the screen the better. If your $30K product instantly becomes much easier to sell when you replace buttons with touchscreens without increasing the costs wouldn't you do that? I guess you need to have a niche, snobby traditionalist brand to be able to reject that demand from the consumers.
Does this ever happen? I've never heard anyone express jealousy regarding not having a big enough touchscreen in their car. I've heard several owners of modern cars with touchscreens bemoan how complicated and slow to use they are. In my experience literally no-one who actually buys and drives cars thinks they are a good idea and many people - including myself - are deterred from buying a new model specifically because of the technology.
The thing is, it's actually really hard to judge quality of a design(takes a bachelor degree in Industrial Design and masters in related field and a few studies like the one in question to objectively evaluate a design). Most people like the new trendy one and unfortunately in cars that's a large touchscreen.
Don't think of car enthusiast, think people who like the car because of the shade of its color and feel of the leather - which is most people.
Those are exactly the people I'm talking about though. I'm in the UK - maybe the current culture is different here to some other places?
Of course it's also possible that my own experience hasn't been representative but I've heard the same story so many times for so long now that it's hard to believe I've encountered some freak sample of outliers.
Also, chalk me up as someone who has never heard a positive thing about car touch screens after a week or so of interaction.
Horse hockey. Spend a couple years in Quality Assurance with your eyes open. It:s trivial to seperate wheat from chaff. The key that your Industrial Design might give you insight on is the fact that Industry has decided unilaterally that cost to produce > joy of end user in use. I.e. if it's cheaper to make and sell, it's higher Quality, rather than it's damn good, now lets streamline it.
Yes, your process weighs into it, but I assure you, the cognitive load of a haptic interface vs a touchscreen is so much lower it's absurd to even try to compare. If you really care about the end user, you take the time to get them buttons, and don't distract them with touchscreen finicky BS.
Precisely. Millions of people (intelligent, rational, highly-educated) still buy cars with specific color/trim as their primary motivator. Until the trend reverses, a screen will continue to be a value-add to any vehicle because of the "modern" association.
Honda has features they implemented in some attempt to streamline the experience but you still get lost easily.
After having a few vehicles with large touchscreens and then buying an F-150 XL to simplify, I can't even describe the elated feeling of operating a vehicle where the screen does what it's supposed to do with the vital controls all being physical. Yeah, I look like peasant but I get to keep my sanity.
But maybe the jealousy is a too strong of a word.
Relative to other iPhones, yes. I read that it accounts for 3% of iphone 13 (Pro, Pro Max, mini, standard) sales. The 13 line itself accounts for about 75% of sales. If Apple sold 40M phones per quarter, that 120M of the 13 line, so 3.6M of the 13 Mini. At 699, that's a 2.5BN business. Not too shabby.
Actually, Porsche Taycan apparently has an amazing knob. MKBHD was very impressed by it[0].
[0] https://youtu.be/BAZX9p2oGOg?t=631
Back in the day, you could easily tell the difference between an expensive high-quality amplifier and a molded piece-of-plastic mass-produced boombox.
Car companies are already kind of doing this, usually just a button or two on the steering wheel. But IMO the entire dash should be a bunch of blank configurable buttons.
You really don't, and the fact that people consider it a given that you do says some very bad things about society. You should be buying the things that work the best for _you_, not the ones that will impress your friends.
I mean, fsck them. If I had people in my life who though like that (I don't), I'd get rid of them. If they're family and cannot be simply cut off, I'd minimise the contact.
I much prefer to just stick to like-minded people and not try to be friends with people where our fundamentals are completely at odds.
I already have a screen showing what radio station is playing, and one on the dash telling me where to go. If a screen is required for a backup camera, just combine it. If I'm reversing I probably don't need the satnav anyway, whereas if I'm reversing or using satnav I probably do need other functions which just means you need an even bigger screen so you can fit everything on.
There is so much opportunity for better regulation, without making more numerous regulations.
But that's ok with touchscreens.
Do you have any evidence for this? Seems pretty outlandish to me.
> 1. It's cheaper to produce; 2. It looks more expensive, so the price goes up; 3. Testing audiences respond positively to shiny lights; 4. Fossils decided that this is what the young people want.
What does 4 even mean? If we took 1-3 as fact, then should businesses have disregarded them and instead made something more expensive to produce that looked cheaper and sold for less because people don't respond so positively?
Mazda also has this beautiful dial-joystick which we can operate in a rested position. It is so intuitive that I stopped using the touchscreen console itself. On the other hand even when we operate using a dial and buttons we take our eyes for an instant to look at the screen to check the changes. Now imagine looking away at a touchscreen just to see what operation to perform etc. This is a major distraction.
I barely use the touchscreen and when not needed I outright turn it off. This is quite easy because BMW has 8 buttons that can be mapped to any function in the touchscreen including turning off the main screen.
Another thing I enjoy is the gesture detector. It sometimes has false positives when I gesticulate a lot but it works when I actually intend it to. It is very satisfying to mute the radio or change an annoying music with a hand gesture. If they would keep trying to integrate and perfect it I think it would be the right direction for innovation.
Touchscreens are fine when parked or for the passenger. Anything else they are useless and often have too much distracting info, so they are turned off.
Was looking at [1] and even with the joystick it is less than ideal. I would need to try though. The joystick is good to navigate the screen but navigating a screen while driving is still not the way it should be.
[1] https://www.youtube.com/watch?v=dPWWbvcs4Dw
In a previous discussion someone mentioned that part of this trend toward screens in cars is that new cars are now required to have rear-view camera. So once you are required to have the screen, it's really almost nothing to waltz over to touch screens. Of note: the "winning" car is so old it doesn't have rear-view cameras.
My MB is a UX disaster.
Have a guess how many controls there are in the car for navigating the (non-touch!) screen?
1? Nope. 2? Nope. 3? Yes 3. A touch surface in the centre console, a spinning wheel in the centre console (which is also a joystick), and finally a little joystick thing on the steering wheel.
Volume controllers? 2.
And don't get me started on how dangerously absurd it is trying to switch between MB's own system and Apple Carplay/Google Auto whilst driving.
Is there evidence it was 60+ year-olds who decided to lean on touchscreens for cars?
If that's just an assumption, isn't an equally likely ageist guess that it was pushed by people who came up through the ranks in the era of "UX"? (Since I'd expect that old-school, pre-UX human factors engineers, who grew up on research coming from aircraft cockpit optimization, safety, and UI in service of the user... would research the heck out of a new technology option like this.)
Tesla does not skew 60+ anywhere in the company, and they introduced these oversized screen based displays years ago.
So on you four bullets above:
1) True 2) I don't know, perhaps? 3) Maybe a quick 'image' audience, but are they doing usability testing? 4) Completely false.
The big weight is on point #1, for two reasons.
1) Those displays may seem expensive, until you actually price out the panels they are using. Then go and see what those physical buttons cost. They are not cheap. And there are a lot of them. And both technologies have micro processors behind them, so using physical knobs and buttons doesn't save money there.
2) Using modal displays to cover multiple controls saves dashboard real estate, and eases design constraints. Designers love it.
One of the things I hate the most, is that I want a mostly dark interior when I drive at night, and now I'll be stuck staring at an illuminated display that I hate using in any case.
Another of the many reasons to decry the death os Saab as a car company.
Later edit: Added link to YT video demonstrating Saab's night mode [1]
[1] https://www.youtube.com/watch?v=Xgh2zbifn7E
You know they're not. If they were, nobody would ever replace a knob with a touchscreen.
* The one and only physical button car took 10 seconds total to complete their tasks
* Two touch-screen cars (Volvo C40 and Dacia Sandero) took only 13 seconds to complete the tasks
* Most touch-screen cars take 20-40 seconds
These results are certainly consistent with the hypothesis, "A moderately well-designed physical interface is likely to be better than an extremely well-designed touch-screen interface". But it's not really enough data to support the hypothesis that all physical interfaces are better than all touch-screen interfaces. You'd want to see what the curve looks like -- with it so close, it's quite possible that some, or even many, physical interfaces would take longer than 13 seconds for their benchmark.
And if you slow people down by 30% but reduce costs by a significant fraction, I think that's probably worth it.
I suspect that the engineers fighting them is really just a case of the hardware team and the software team not understanding the world the other lives in. The hardware team is working with a slow as molasses processor that is the only thing thats been approved for the ridiculously rugged life that a car CPU lives and the software people don't understand that just because a webkit rendering engine is completely fluid on their 6 month old Precision workstation it won't be on a 500mhz in dash processor.
Doesn't stop Toyota for wanting a solid $1000 to replace the display in my 2014 Corolla. Someone's pocketing a lot of money.
This makes a ton of sense for displaying state.
For manipulating state I need tactile physical controls.
This is how computers work, and for good reason. I have a big screen to show state, and keyboard + mouse to manipulate it.
It all makes sense from just the financial point of view. So that means it isn't going away any time soon, unless there's a huge backlash from consumers.
Perhaps the best thing we can hope for is 1 car manufacturer deciding: "Buttons first, touch screen(s) second."
Let consumers decide with their wallets. Though, I wouldn't be surprised that many consumers go for an inferior product just because it looks cool. Because that, unfortunately, is how humans work.
Sort of related, I have the exact same issue with portable music players while walking or cycling. Most of the time the only task I need to do is play/pause or forward/backward track.
For a player with buttons it takes a small amount of attempts and after that you've learned the position of the buttons by heart and can control the device even while it's in your pocket, without needing to see it. Usually aided by some tactile feedback. Fast, convenient, and somewhat safer since we're talking traffic situations.
With a touchscreen-only player that is much harder, sometimes impossible (depending on which screen you're in the controls might not be in the same place or not be there at all).
Sad thing is, this was already the case like a decade ago, leaving me wondering if designers have any pride in their UX, simply don't know they're doing it wrong, willingly just focus on other things apart from usability, etc. In any case: driving a heavy vehicle at high speeds should be the last case where simple things like switching a radio station actually requires you to take your eyes of the road. That's just insane.
My steering wheel has volume controls and the environmental controls are still button based.
The updates are only for bug fixing.Maybe this will change with SaaS but today i never heard of any Car company which does this. And no, i don't consider Tesla a car company.
Please don’t play this game where a bad design decision was finally recognized as bad design and a scapegoat is found rather than admitting the “experts” who did it have no clothes.
UX branding itself “UX” rather than any of the half dozen other names we used to use was a clear statement of “It will be different this time, I promise.” It wasn’t. The design trends we got were different, but bad interaction design is still bad interaction design.
I do agree with the premise that physical buttons and knobs are generally far superior to touchscreen UI's, at least for the common core basic things.
However, I don't agree that it's about the boomers, or the capitalists, or any other Internet strawman forcing something onto the masses against its wishes. I think it REALLY IS a matter of test audiences and "casuals" having tastes that differ from power users and other people that think deeply about a thing. You see this in many different domains.
Hoping it comes back after being turned off overnight, or I’ll have to find a manual online somewhere.
I don't agree with not giving you a paper copy of the manual though, especially when its a luxury car. The digital manual should be in addition to the paper one. They shouldn't be pinching pennies like that.
One of the features of modern cars is conversational UIs (CUIs) that enable a truly hands-off approach from controls altogether, both touchscreen and analog.
I hardly use my car's touchscreen except for discoverability, and definitely not while driving. Speaking to the car is easier and more efficient:
- "I'm cold"
- "Turn on the windshield wipers"
- "Navigate to work"
CUIs have evolved to the point where controls are generally a hindrance to consumers and manufacturers both. Why include expensive, breakable analog controls when you can give the driver a better user experience hands off.
Buttons are the way. Or those cool switches from fighter jets... I'd like to see those in a car.
I can't speak to all cars, but this isn't the case for Tesla. It's a very good user experience. It may be inconsistent across manufacturers right now, but as that evens out I don't see a barrier to more adoption.
I don't agree - I've been very happy with my car's CUI. I can't see going back to analog controls or touchscreen. There just aren't that many commands that I need to execute while driving, and the CUI understands them all.
"Can't open the trunk with the car in movement."
Clicking a button is orders of magnitude faster than casting a spell. Voice controls fail more often than analog controls break. You do not want a delay to turn on your wipers when you are at 60mph and hit a sudden localized rainstorm.
It depends on how many buttons there are and how familiar the driver is with them. I like not having to remember/look for the positions of analog buttons on the dash.
> Voice controls fail more often than analog controls break.
I haven't had a voice control failure, but any system can be designed with varying degrees of reliability. CUIs have a lot to offer in the way of safety and convenience.
If it's your car, this problem will solve itself for you in short order. There's not that many functions that a dashboard needs to do that you won't familiarize yourself with it in a month or two of driving.
Look to a computer for an example, some shortcuts that you probably use on your keyboard are downright arcane, but because you use it so frequently it's probably natural.
- you never drive with your windows open?
- you never drive with passengers that you talk to?
- you never listen to music in the car?
"I'm cold" well if you say 'goodnight' to Alexa, she says goodnight and carries on with the music which isn't what I want. I can imagine "I'm cold" being the same, and even if the car does get the hint what do they set the temp to?
"Navigate to work" I know where my work is, I mostly want to navigate to places I don't know, and are therefore not in the memory. So it's "car navigate to some street, city X" "Would you like 1 some street, 2 some street, 2a some street...."
Or "Do you mean some street or sum street or summ street?"
And that's assuming they know the pronunciation, the locals of Slaithwaite can't agree on the pronunciation. And theres many places pronounced weirdly.
If I have to press a button to make it work I might as well press the button that does what I want.
It seems like a good idea for adjusting the climate control but I apparently didn't pay enough and it replies something along the lines of "I'm sorry Dave, I'm afraid I can't do that."
So far it hasn't offered a monthly subscription to enable that feature.
Honestly I barely feel the need to use the touch screen in my car while driving, and if so I always make sure autopilot/high way assistant is turned on as extra safety.
Familiarity is my personal favourite part of driving. Knowing the road, how much input you need to apply to a turn, knowing where the buttons for things are, just being able to feel for a control and know its purpose. All while my eyes are on the road.
You lose this with a screen.
Simply put, with physical controls you can operate them without looking. That is impossible with a touch screen.
And the new generation are all using touchscreens which need you you take eyes off the road to find the fan settings (return to home screen, touch near the bottom centre, then find the wedge-shaped fan speed widget and adjust that). Temperature needs to pop up another pane and set that by clicking the temperature number and then manipulating a bar chart thing. None of this except the "return to home" button is tactile in any way.
Before: turn the dial. Done. Temperature is the one next to it.
Adjusting the sound balance is downright dangerous (pull down the Android-esque menu and click though levels in the UI). Old car: press the tactile centre of the volume wheel until it says balance, then use the wheel.
If we had to dial every number we want to call manually, physical buttons on phones would outperform touchscreens too. But nobody does that anymore. And it is nicer to select the name of a friend on a touchscreen than to type it with physical keys.
Same goes for cars. Humans spending their time keeping the car in lane, stopping at red lights, starting when the light turns green, doing turns etc will phase out more and more over the next 10, 20 years.
The interaction we still want to have with a car is probably nicer on a touchscreen. Especially when you are not dabbling with a steering wheel anymore.
Stuff like seeing the route on a map, selecting waypoints etc. We would not want physical keys for that when we are at home at our computer, right? I think dedicated keys are just legacy from the "I'm busy with the steering wheel and need to do other stuff blindly" area.
Much of it will probably also move to voice control.
IF we get to more autonomous driving, sure add in the touchscreens, but until then I think we may have started that trend about a generation too soon.
Google started development 13 years ago and now already has fully autonomous taxis on the road in a 4 cities. If they double the number of cities every year, in 10 years they will be in over 4000 cities.
Tesla started around the same time. Looking at videos of tesla in self driving mode, my feeling is that without human interaction, it would crash maybe once every 50 hours of driving. Double that every year and in 10 years it will crash once every 50000 hours. A human driver, driving 1 hour a day, would need 136 years of crash free driving to achieve that. But the average driver has 4 accidents in their lifetime. So we are already in superhuman territory by then.
While there has been a lot of progress made to date, much of it (IMO) has been around the low-hanging fruit kind of stuff. Sure, we've mapped a lot of roads and covered many of the basics. But none of the systems are really able to do human-level context awareness, like you see a bunch of kids running around near, but not on, the road a few hundred feet up, might want to slow down or at least be extra aware, or people clearly driving in "tourist mode", etc.
My personal assessment is that we are still 20+ years away from the point where the human driver is mostly along for the ride (and thus to the point of this thread, can fiddle with a poorly implemented touchscreen UI without risk). I am led to believe the ultimate solution is going to involve altering/enhancing the road infrastructure, along with the general improvements for the in-vehicle stuff (which itself has at least a decade of development still to go). Combine those things together, and we're still quite a ways off from the vision that was sold 10 years ago.
I expected something like this. That's why I said "quantifiable". We humans are animals of habit and have a hard time imagining a changing world. But every time a few decades pass, we look back and see - damn! - a lot has changed.
A way to get away from feeling and towards a rational prediction is to look at rate of change. In my experience, doubling the performance of a new technology (self-driving is only 13 years old) is usually doable. And that means 1000x improvement in 10 years. And from where we are now, that gives us Waymo in every major city and Tesla with super human capabilities.
I do not think your rate of change assessment is accurate. With many newer technologies, like the DNN/CNN frameworks, we see a more or less immediate order of magnitude improvement, and then declining rate of improvements as the easier bits are done and we are forced to do deeper development or refinement.
Saying that Tesla has super human capabilities sounds to me like you are not assessing it rationally. Super human would mean, to me, that it manages to do better than humans in extreme conditions. Instead, we have seen a number of very preventable causalities in the current generation systems.
In terms of self-driving, some factors are:
The amount of data grows exponential. Not only do the existing Teslas keep adding data, but more and more Teslas are added to the fleet, accelerating the pace at which data is generated.
Crunching the numbers gets exponentially faster, because compute power is growing exponentially.
The algorithms used to crunch the data become better.
As more and more revenue comes in, more and more can be spent on data crunching.
The sensors become better.
Maps become better.
All of these factors multiply. Adding another exponential force. Even if the data would grow at a linear pace and the algorithms would get better at a linear pace, this would result in exponential improvements, as these two factors multiply.
It really isn't for me, it's much easier with the T9-style 3 letters per key on a physical keypad and then up/down on the steering wheel.
Some cars have a click wheel that lets you set letters one at a time, which also works well.
What doesn't work well at all is typing with one hand on a centre console touchscreen.
https://youtube.com/watch?v=9BnLbv6QYcA
Not so good for typing, but can be quite good when there's a good UI that narrows down options for you at each entered letter.
I wanted to make a call to my partner recently, whilst driving. I have an old car, with all physical controls. But my phone was in front of me with maps running. I said "OK Google.." and waited for it chime that it registered. "Call [my partner's name]", I commanded it. It responded that the name was unrecognised - even though it repeated the name back to me, so it wasn't a case of mishearing. I repeated this at least 5 times, getting increasingly frustrated.
Then eventually it started calling. Joy! But I notice it doesn't say my partners name on the screen, just a phone number. "OK", I think to myself, "perhaps this is some quirk of voice dialing, it doesn't show the name" (I don't know my partners number by heart so I could not recognise it).
It rang through to voicemail, and whilst the name given by the voicemail recording was correct it was definitely not my partner's voice. It turns out my phone had just dialled a random mobile number from the Internet that matched that name. It may have been a business, or a personal number, I'm not sure. Either way, I would never want this behavior!
I guess the point is that voice commands have terrible error handling and recovery modes. Give me physical controls any day. Even if they're slower, at least they are accurate, discoverable, and do not make guesses about intention