> If you need to take your eyes away from the road, it is a shit ui.
I'd add one more point specific to vehicle (or other safety-related) applications: if it keeps changing, it's shit. I shouldn't have to hunt around for where that control went today any more than I should have to hunt around for where it went the first time. That is why I think touchscreens with tactile feedback don't solve this problem, even though they're cool in general. That customizability is worthless because the controls shouldn't be changing anyway. Just put in real buttons, knobs, and so on.
Can't comment on this specific cost, but ~20 years ago when I worked at Ford I remember seeing a little poster on the wall celebrating how they value engineered $200 or so out of a Taurus by replacing or eliminating a lot of small parts and pieces.
Tesla is bumping up against 1M cars built per year. Saving $10/car can add up to real money.
Depends. SKU count is lower for sure. Wire count is less too. And in Tesla's case (Model S) where there is no real dashboard with a speedometer at all - it really would make a difference.
Didn't found how much Tesla pays for it's screen, but at $300 and even $500 I wouldn't be surprised if it would be still cheaper.
Touchscreens (or touch surfaces) should not be used for any kind of vehicle function. period.
If you need to take your eyes away from the road, it is a shit ui.
If you can't feel but have to look to see the result of your interaction, it is a shit design.
If you need to look at the center console and down, it is even worse.
There is only one reason to get rid of all the buttons and switches, and that is design. because "ape goes ohhhh!".
And that should not be a consideration that is prioritized over safety.