66

Last week this video about a new interface for touchscreens in cars popped up and created a reasonable buzz, maybe you've seen it already. I just got to watching it and it got me thinking about the problem too.

Most new cars all come with touchscreen for radio's with extra features like navigation and traffic reports. The problem is car manufacturers cram too many features in too little of a space and you can't possibly control them without endangering yourself by taking your eyes of the road for a certain amount of time. Older car radios you could control without looking because of the feedback you'd receive from the buttons through your fingertips.

The guy in the video has made a more intuitive interface, but it has a steep learning curve.

That got me thinking: what advantage does a touchscreen have over good 'ld buttons.

It has some visual advantages for navigation and maybe album covers, but what about temperature and volume. To my opinion it's easier to remember the location of certain knobs than to remember the right finger placement on a touchscreen. And what feedback would you have? For example: I would know for every click the temperature will go up half a centigrade. For touchscreens I would have to look at the screen to see what temperature it is at. With temperature you don't have the instant feedback you get with something like volume. Also, buttons still have icons on it. Can't remember where the button is, look quickly. Can't remember what finger combination to use? Now what? You're screwed.

Is it just a flashy technology? Or do touchscreens have advantages that make them capable of taking over the entire car console?

Update
Apple introduced their car interface system in collaboration with Volvo.

14
  • 14
    Personally, I hate touch screens on phones and on cars for the same reason you mention: you can't use them without looking at them. That doesn't stop me from driving my Prius, and it wouldn't stop me from buying a Tesla (if I had the money...), but I would still prefer tactile feedback so I could 'look' for buttons with my fingertips.
    – Patrick M
    Feb 26, 2014 at 14:36
  • 12
    @PatrickM I used to be able to text with my hand with the phone in it, in my jacket pocket, while on my bike. Feb 26, 2014 at 15:03
  • 9
    That's the kind of valuable life skills that today's youth will never learn. Good grief, complaining about this makes me feel old :-p
    – Patrick M
    Feb 26, 2014 at 15:08
  • While this is interesting, is there really a question and answer here, or are we all just happily ranting together? Feb 26, 2014 at 23:49
  • 4
    @Paul Are you really texting in your pocket or are you just pleased to see us? Feb 27, 2014 at 23:12

11 Answers 11

52

We are really at the early days of touchscreen technology.

While audio feedback and advances in haptic feedback could make this slightly more viable, I see touchscreens as an interim workaround on the journey towards gestural (+ audio) input.

The mistake in this design shown in the video is (in my opinion) using a touchscreen at all - i.e. a 'touch sensitive surface' that is also a screen.

Touchscreens are great for computing devices or a thermostat or even for passengers in a vehicle, but not for a driver's in-car interface. In a car, the driver's eyes need to be on the road. Therefore we should be thinking about separating input and visual output completely.

I would be happier if designers were using a touch pad (no display) or gestural spatial input combined with a dashboard based screen (transparent or reflected) being used for the purely visual feedback and positioned in the lower windscreen line of sight. That way, there is not even the temptation to look at what your hand is doing or watch the pretty circles and the animated sidebars - because there's nothing to see.

So not just flashy technology, but misguided technology.

However, I have one more issue. I foresee a meeting of technologies where by the time a well designed, tested, approved system like I describe above can come into place, the need for it will be eradicated by an influx of (mostly) automated driverless cars where it doesn't matter if the driver's looking at what they are doing anyway.

10
  • I like the idea of just the touch pad and a transparant display somewhere on the windscreen. Feb 26, 2014 at 13:19
  • Actually I just noticed that the guy who made the video already suggested on twitter about a head up display for current state (says he didn't want to get lost in details on the video), but not seen mention of a touch pad vs touch screen Feb 26, 2014 at 15:41
  • 5
    @Paul If there is a screen, you will look. There has to be no screen for it to work for the driver. Feb 26, 2014 at 15:56
  • 15
    +1 In a car, the driver's eyes need to be on the road. A million times this; it's exactly why texting while driving has been made illegal in many places.
    – SpellingD
    Feb 26, 2014 at 20:11
  • 2
    @Paul perhaps, and that looks like a cool thing to have for a car, but typically there aren't nearly as many jets in the sky in a common location as there are cars on a freeway. Focusing on anything other than the road presents a risk, even if that thing is in the path of the road.
    – MDMoore313
    Feb 27, 2014 at 17:49
34

Thought I'd throw my two penneth in as a former Automotive Interaction Designer for a large British car manufacturer in the premium and off-road/footballer market owned by an even larger Indian company. Starts with "Jag", ends with "...nd rover"

Anyhow, for those of you familiar with those brands you'll know they use touchscreens. I'm not a fan. The NHSTA guidelines (and they are guidelines, not 'rules') are the most researched standards in infotainment, and they are very strict on eye-off-the-road time.

Touch-screens, IMO, require constant visual feedback. Especially the kind on display here. When you're driving that's not something you can easily and safely commit to.

So, that's the safety aspect covered - now lets look at it from another angle. The video shows the system in use in a very manicured environment away from the variables we encounter in cars and whilst on the road. So, the differences are: - The user is stood directly in front of the unit - The unit is completely static - The user is ONLY using the unit

Standing directly in front of the unit changes the usability dramatically. In infotainment HMI design there's easy and hard to reach zones. If you're sat to the left of the unit (if you drive on the right as most do) the top right corner of the UI is a sort of dead zone.

The unit being completely static makes precision much easier. I'd be interested to see this used on a rig which shakes it round as much as it would be in a car.

The unit is being used by somebody who isn't driving in the video, meaning the ignorance of familiar paradigms has less of a discomforting effect to the user than it would to one who was trying to drive at the same time. Driving requires a lot of attention and should be your priority, not fiddling with your iPad to change the volume to 12 because you're on 11 and odd numbers freak you out (it's amazing how common that is). Using the temperature example lets compare a click wheel with tactile feedback (as seen in an audi) with this system. Your car is cold and you want to set your climate control to 25 degrees.

Tactile click-wheel: You can see on the display that your CC is set to 18 degrees and you want 25. You already know that a click on the wheel is half a degree, so you need 14 clicks. You put your eves on the road, go a few clicks up, check what you've set it to and adjust as required.

This system: You activate temperature modification and swipe up. You have no idea how far up because the distance of swipe isn't relative to the screen any more. Looking over at the screen should reveal the volume but as your eyes were on the road before the screen has now gone blank.

So, safety and environment covered, let's look at the user.

Car users are not tablet users. They're older (average), have a higher chance of being "technophobes" and are also more keen on getting the job done easily compared to a tablet which is often a leisure device. By introducing a system like this we do away with almost all automotive paradigms that have been established over the last 50 years, and your average user won't like that. We also introduce something that is, IMO, gratuitously modern - installing fear in many.

Maybe one day automotive UIs will be like this. I'd also like to point out how keen I am on the system from a visual angle, but I think this has been designed with mobile paradigms, environments and users in mind and isn't fit for purpose. If this was a home automation system it would blow all others out of the water, but not in a car.

I'd love to work with this guy and find ways to make it work as a user experience overall.

8
  • 6
    P.S. not everybody has 3 fingers...
    – TJH
    Feb 26, 2014 at 12:46
  • @TJD Agreed on all points. The curated video with calm 'music' and voice also very much struck me as inappropriate considering real world target environment. I've worked in fields of virtual reality, visualisation, CAD, UI and UX and would also like to work with this guy :-) Feb 26, 2014 at 12:53
  • Thanks @RogerAttrill - I've got an amazing idea in my head now to film it in a zombie apocalypse setting using a Merc G-Wagon :) Gunshots instead of calm music! :)
    – TJH
    Feb 26, 2014 at 12:58
  • 1
    @TJH what would your ultimate solution/car-ui be? As a former interaction designer in that industry you MUST have tons of ideas. Feb 26, 2014 at 13:31
  • 2
    @Paul - completely agree with your summary there. Automotive looks at the consumer mobile device world with it's cool product videos and cult following and thinks "I want to be like that" - the Audi system in my previous car was completely non-touchscreen. It had a clickwheel and a set of contextual buttons whose values were on screen. Perfect! Your point about looking away is spot on. "Hey TJH, hows your Tesla?" "Oh actually, Paul, I crashed it and lost my legs because I was using a touch-screen web browser at 100mph! Oops!"
    – TJH
    Feb 26, 2014 at 15:08
6

As others have said, a touch screen is an inappropriate interface for a driver. The advantage it has over buttons is cost and perceived flexibility. The cost of an assembled panel of switches and knobs is substantially higher than installing a single touch screen. And the more items you want the driver to be able to control, the more knobs and switches you have to install - and there's a limit. Too many identical controls can be confusing.

The reason I say "perceived" flexibility is that when people see a touch screen, they imagine it can do more than it already does. However, cars are traditionally very static devices: people get used to "here's the way I turn up cabin heat" or "here's the way I change volume on the sound system". There isn't much incentive to add or modify those interfaces, because your customers will be mad if you change it from what they've grown dependent upon. Even if you could think of an extra thing you want your drivers to have to do (instead of driving), it would be difficult to deploy.

What's almost as bad as a touch screen are non-tactile touch-based controls. At least one automaker has capacitive touch sensing buttons for their controls. An experienced driver knows his vehicle, such as "grab blindly for fan speed knob, feel right two spots, rotate knob clockwise for defroster, back one spot for floor heat." They use the first reach to locate a frame of reference, then count over to reach the controls they want. If you have capacitive controls, it's easy to trigger unwanted actions just by touching the panel.

Even worse are joysticks and/or knobs that accept vertical motions. A car bouncing along a rough road will cause the user to incorrectly trigger undesirable outcomes. I once had a car with a vertical slider for heat controls, and it was very difficult to set it to anything but "full heat" or "full cool". If you wanted marginally hot air, a little bump could easily send you full heat. Rotating knobs with clicking indentations seem to work well, as do toggle type switches. Sideways sliders have given me less trouble on the road. If you can have a separate stud to serve as a reference touchpoint, mounted next to the display such that I could rest a finger on the peg while attempting to move the slider, that would help me with fine adjustments.

Right now it seems that touch screens are installed in cars only for the sake of giving a user's dashboard a "modern" look and feel. Instead of benefits, they primarily offer distractions. And it's obvious that whoever built the UX did no behind the wheel studies.

5

I don't know this for sure, but I suspect that manufacturers are at least partly embracing touch-screens for business reasons that have nothing to do with UI, or in fact the user at all:

Touchscreens are software controlled.

That means that you can get the rest of the dashboard designed and put to bed, parts ordered, molds made, etc, while continuing to tweak the look and feel and operation of the touchscreen almost up to when you go into production.

Like many things in today's world, using software to do something decouples parts of your design process, allowing for more overlap and (in theory at least) better time to market with snazzier features. This isn't necessarily a 100% win for the end user, but it may be the best thing for the business, and in some cases (perhaps not this one) it may be better for the user overall.

4

In my view this is a clear example of UI attempting to lead the User Experience. If users are using a touch screen than this kind of interaction is useful. Currently the touch screen experience in most cars, especially the Tesla, is terrible, ugly and less usable than physical buttons.

In this case the user has to learn what all the different gestures are. Any interface that requires learning will have lower usability unless the user is focused on it for a repetitive task. With a car interface they won't be focused on it!

You're right - it's a great interface idea but a terrible user experience.

3

I agree with all guys that visual feedback should be eliminated. It's very risky to check changes on the tablet's screen because you have to move your eyes out of the road.

1) Volume control

It's much easier to change the volume with buttons on the steering wheel. Because I don't need to move my hand away from my wheel. And to do the required gesture on the tablet.

And I don't need to set the specific dB ) I just need to make it louder OR quieter from the current value.

To click the button much faster then to make the suggested gesture.

BTW, I get the immediate feedback while changing the volume. And I don't need to see the current dBs.

I'm totally satisfied with my steering wheel' volume control' buttons.

2) Temperature

The ideal solution IMO: car automatically detects my body temperature and set the car temperature on the basis of it. But of course it also should take other factors into consideration: the temperature on the street for instance, so if it's cold and this is the first user's visit to the car then it should maximize the temperature in order to warm the user and the car.

And if the suggested temperature doesn't suit me then I can make it more hot or more cold, but I don't need to know exact degrees. I just click on physical (+ or -) button and feel the effect in some minutes, because it's not possible to apply new values immediately.

Nowadays you must guess the degree to set - you set it to one value, then change it to other value, then again and again.

Let the car study your preferences automatically. And it can even study your behaviour - when you change the temperature in different circumstances. And apply this knowledge by setting the temperature with ever-increasing accuracy.

But this is far into the future )

2
  • I like the solutions, but unfortunately it's for a future setting. On the same note you can also talk about implementing a Siri, Cortana or Samantha. Better voice recognition. It's rumored Apple and Google are already working with automotive companies. But I love the idea of just telling I can't see anything through my windshield and the front heating comes on. Feb 26, 2014 at 15:52
  • I really really hate car ac/heater systems were you set a target temperature and it tries to figure out how hot or cold to blow out of the vents. I much prefer the older style where you set the temperature you wanted the air to blow at. There are several hot/cold zones within a vehicle and no good place to put a temp sensor.
    – NotMe
    May 17, 2016 at 23:05
3

Exhibit A - factory entertainment systems

  • physical (often single-function) buttons
  • monochrome/minimal displays
  • shallow menus

Exhibit B - more complex aftermarket systems

  • greater visual emphasis
  • deeply nested menus
  • more functions

My hunch is that manufacturers noticed people replacing A with B and decided to offer B by default.

4
  • 6
    But the general car-buying public is a much different group of people than those who put in aftermarket items, and I think they have a more mild taste
    – user70585
    Feb 27, 2014 at 4:47
  • 1
    "People don't know what they want until you show it to them" - Steve Jobs. People are a poor judge of what they need, it's often different from what they want. If you're correct, this is an excellent example. Feb 27, 2014 at 7:56
  • Good point @Dar - there's a big difference in the target market for after market systems. People only opt-in to those, so whilst some (you or I) might find them easier to use and a better user experience others (our Grandparents) might find them more complex. This is fine, as the manufacturers are simply choosing to not sell in to those markets. In the scenario one cannot decide on the infotainment unit all tastes must be catered for. Paul raises a good point and the quote is spot on. Solution? Maybe we can be offered opt. B software on our opt. A units?
    – TJH
    Feb 27, 2014 at 9:03
  • 1
    @Dar agreed, and it's unfortunate that car makers would follow the aftermarket stereo trends, because the UI on those is typically worse in every way compared to the stock system. Feb 27, 2014 at 17:40
2

Putting so many fingers in so small of a space seems like a pain in the neck, and more importantly, the UI is in no way whatsoever self-explanatory. To expect everybody to learn these controls comfortably is ridiculous. That, and the finger closeness is counter-intuitive as shown; the two-finger together and apart controls are not terribly related to one another; definitely not enough to feel natural.

2
  • I totally agree. What solution would you propose? Feb 27, 2014 at 7:48
  • I would probably reinstate the tactile controls of days past. Though a siri-esque voice control may work too; I think it will be too distracting. For a few reasons actually: Initiating the control might be distracting if placed poorly (if it has no initiation, it could be very, very annoying), and dialogue has been shown to take focus completely off of the road, which is a problem to which I don't see a solution in a voice-based control system
    – user70585
    Feb 27, 2014 at 15:03
2

The obvious solution is to replace the entire interface for non safetly critical items with voice control.

"Computer: Put on Radio 4."

"Computer: Make it 2 notches louder"

And eventually if you get bored driving on your own you'll be able to chat to your friendly in-car entertainment system....

2

Touchscreens offer almost infinite possibilities for variable user interfaces, but tactile buttons offer accuracy, speed, motor memory and safer operation in a moving vehicle. There seems to be a an almost stalemate between the desire to offer a flexible UI and the need for physical buttons in a moving vehicle. But there may be a solution...

Why not offer both? If there was a way to offer a low-cost commodity TFT display, and cover sections of that display with tactile keys, rotary controls and buttons, but bring mapped sections of the display up into those controls using bonded fibre optics, that was easy to manufacture, low cost and offer high aesthetic appeal, surely this would be the perfect compromise?

Previous attempts to use bonded fibre image conduits were thwarted by the expense of the glass fibre optics, but recent developments in polymer technology have yielded excellent results, making this method of offering Picture Keys for a more affordable and viable price.

These buttons can be almost any shape or size, and the same screen used to label the keys depending on the mode of operation (climate control, entertainment, telephone, GPS, etc) can also be used to display relevant information in an exposed region not covered by controls.

The technology is available now and has been proven to be successful.

0

I personally don't care for the touch screens at all. Having been an accident investigator, I can see the as quite a distraction.

I have driven a number of cars with the touch screen feature and they are a distraction. What really gripes me is that most manufacturers are making them standard equipment and not giving the buyer a choice.

We usually get a new car every few years and now have effectively been taken out of the market for a new car as I refuse to pay for technology that I don't need or want.

1
  • The question is specific around "what advantages do touchscreens have", but you have shaped your answer around the problems with them. That's not what was being asked.
    – JonW
    May 18, 2016 at 8:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.