What Tesla’s Model 3 UI Reveals About Its Vision for the Future | Hacker News Reader

What Tesla’s Model 3 UI Reveals About Its Vision for the Future

blog.figma.com

48 pointstresbonn54 comments

> They weren’t even trying to hide it, the middle screen was basically a giant middle finger to the rest of the car-building world; a not-so-subtle “F-you” to the naysayers, the disbelievers, and the other car companies who insisted on playing it safe.

I don't know if I would call it that.

>I don’t think Tesla did this simply to be edgy, dangerous or different. They did this to prepare people for a world without drivers.

In a world without drivers, a handle to open up the glove compartment, physically-controlled air conditioning vents, and physical HVAC knobs still makes sense.

> Instead, the Model 3 has a 8:5 aspect ratio and a resolution of about 1920x1200px. ... suggests that Tesla made their center console for watching things, not for driving the car.

8:5 is 16:10. What I really think happened is that Tesla took the cheapest available off-the-shelf 16:10 LCDs. Because what the Model 3 really is is about cutting costs to get the cheapest electric car on the market, and ultimately I think that's what informs the Model 3's design the most. A cheap electric car is in itself a noble goal. I think that it's wishful thinking to think otherwise.

This seems to be kind of an ad for Figma. That's fine, but it's a poor ad. It suggests that Figma is a tool that embodies a myopic vision of UI design where only the screen matters.

It very much looks like off the shelf 16:10 touch panel. I've been told that the center display is even mounted to the dashboard with what essentially is 100mm VESA mount.

A decade ago 16:10 was the size we were heading towards, plenty if not most screens were 16:10 for a brief while when we got rid of 4:3. That also transitioned through 14:9 for TV that would work on both screen sizes, it took a while to settle on 16:9 and for everyone to complain about laptops with silly shaped screens.

Imaginably when this project started the instrumentation that evolved into the touch screen was on a 16:10 display, still the common size on Android tablets and sometimes in super high resolution. Or in low resolution - 1680 x 1050 is a size for monitors albeit not so common now.

You don't have to work with computers too long to know that 16:10 is better than 16:9 except for with televisions and dedicated video players. I would expect that given the market for screens and availability that Tesla would go for the correct size, i.e. 16:10, to show full screen video plus some UI buttons.

I am hearing you regarding Figma and also I agree with you on 'myopic vision of UI design where only the screen matters'. It is not how design works.

I am intrigued by the volume for available consumer grad 16:10 vs 16:9. I was under the impression that panel at 16:10 where becoming a rarity. If this is the case, it makes less economical sense from Tesla to have such a screen ratio.

> Because what the Model 3 really is is about cutting costs to get the cheapest electric car on the market

I thought that was the case, but having seen a few in-depth video reviews there still seems to be plenty of fat that they easily could have trimmed without anyone caring. E.g. the charge port cover can be opened electronically; why? It seems pointless to me, since you have to get out of the car to plug the charger in anyway.

One of the stated goals for model 3 is true autonomy... the car will need to charge itself. It will drop you off, go find somewhere to charge.

Self promotion. It's on the outside of the car and may prompt others to glance/notice it and be intrigued by it.

A pretty low cost for the amount of 'increased' perception of value.

To be honest, that dampened my excitement for the model 3 a notch.

In order for that design to make sense in today's model 3, it would mean that very car is just a software update away from autonomy at "we are all watching netflix instead of looking outside" level. And that software update is coming very soon, in a matter of maximum 2-3 years.

I'm skeptical. I can very well see the next Model S, or the refreshed model 3 in 5 years achieve that [1], but really that's sacrificing today Model 3 user for something that's going to never be delivered to them. Not sure I like the approach of "this is a glimpse of the future that is not for you".

I don't think that there is much need for a lot of button, don't get me wrong. To me, just somewhat more feature full lever around the wheel - nothing fancy, just the same you get on most cars and a hud so that the rest of the family can indeed watch Netflix while the driver does not lose completely every driving information. Job done, a taste of the future, but not completely compromised experience for the poor sod that has to drive in the present.

[1] A lot of people that know more than me are skeptical this is even possible short term, but with no money at stake I'm happy to dream.

There could also be a desire to push partial-autonomy features in the short term, when Model 3 owners are making the decision day-by-day whether to drive the car completely manually or whether to rely on whatever level of Autopilot is available now or in 2 years. If driving the car without Autopilot is deliberately made to feel weird and akward, more people will turn on Autopilot in more situations.

Partial autonomy really sucks in less than ideal conditions. Like it's raining like mad in poor visibility, the road is atrocious, the marking is weird, ...

Those are the times you don't have the choice but drive yourself and that's also the time you have to fiddle with a lot of option like finding the right speed for the wipers. Nobody like driving in those conditions, and if Tesla makes the experience even worse that will indeed make people regret Autopilot but maybe not in a good way.

Driving manual in a sunny day in country side twisty roads ? You really don't need anything else than the brake, the wheel and accelerator, no matter the car.

I think this article overstates things a bit. I have a Model S which has a bit more in the way of levers and buttons, but is still really minimal compared to most cars, and it's fine.

The Model 3 is lacking two main physical controls compared to the Model S.

One is the windshield wiper speed control. Both cars have a button to trigger the wipers manually. The S also has five speed settings physically accessible: off, auto-low, auto-high, continuous-low, and continuous-high. The auto wipers are pretty good (this is the old Autopilot hardware, the new Autopilot hardware still doesn't have the software for it) and so I hardly ever have to fiddle with it. I bet a bit of software and hardware improvement can take that "hardly ever" to "not ever" so I don't think the Model 3 will suffer here... once the auto-wipers are actually enabled!

The other is cruise control. The Model S has a separate stalk for that, which you can use to engage, disengage, adjust following distance, and adjust speed. The Model 3 has it on the PRND stalk and it's just on/off. I don't think following distance is a big deal. It's pretty much set it and forget it on my Model S. Speed adjustments are often handy, but a lot of that is because the car isn't totally reliable at figuring out the current speed limit. When it does, manual speed adjustments will be needed rarely.

The Model 3 controls are designed around intelligent driver assist features, but I don't think they're designed around anything close to full autonomy. They're really just designed around good automatic wipers and intelligent cruise control, both of which will probably be ready before most of us mere mortals can obtain one of these cars.

I have a Model S too, and my one irritation with it is that I don't have a physical button or dial to turn off the headlights. Whenever I drive onto a Washington State Ferry at night, it gets me into a bit of trouble.

If you haven't done it; you're required to turn off your headlights the second you board the ferry, there are signs posted, and the ferry operators will often check as you board. One reason is that sometimes car owners leave their lights on, drain their battery, and delay the ferry unloading at the other end. But even as you debark the ferry, you're required to keep your lights off until just off the boat. I think glare and operator safety might be factors too.

But in either direction, if it's night then you need your lights on just before, or just after the boat, and you'll be driving across a narrow little articulated loading dock at the time. Trying to drive in those conditions, and get through 3 taps or however many it is on the console, just isn't going to happen. So twice now, I've had a boat worker stop my car, yell at me, and have everyone wait while I go through the LCD to do it.

That's an interesting problem. I've only had to manually fiddle with my lights once or twice so I hadn't really thought about it.

One thing I think they could really improve would be to enable voice commands for these things. It's still no substitute for a physical control when you really need it, but it would help, and it ought to be something they could add in an update without much trouble.

I have a Model S too. And agree with your first point about the wipers. Not so sure about the second one. Where AP2 is at the moment, I have to constantly adjust speed and distance, especially during heavy traffic. Because I want to reduce the car length during very heavy stop and go and increase it when traffic is flowing. Also, when the speed limit changes, Tesla doesn't bump up the set speed, so I have to manually change it. I love how I can achieve all of these using the dedicated stalk

Following distance adjustment could be made unnecessary with a relatively minor software improvement, where the distance is made more dynamic than it is currently. I'm pretty sure the main reason it doesn't automatically adjust the set speed when the speed limit changes is because it's not sufficiently confident in its knowledge of the change, so it could go away once the system gets better.

With the current state of AP2, the Model 3's controls seem to be a bit lacking, but they should be sufficient once relatively minor improvements come along, no need for anything like full autonomy.

Watching this video: https://www.teslarati.com/doug-demuro-tesla-model-3-review-c...

It seems like there are a lot of questionable UI decisions that would make me hate driving it after a week.

Opening the glovebox with the touchscreen is the epitome of bad design, IMO. After watching Doug demuros review of all of the motorized components of the model 3 (the doors opening, the gas cap, the glove box, the ventilation direction, and probably more I've missed) I see a car that's going to break often. keep it simple stupid.

Not only will things break, but they'll also be a pain to use. People will hate doing simple things in their car, and this will make them like the car much less.

I recently drove a Cadillac XTS which has the motorised push-button glove box. It didn’t really add that much convenience over a traditional pull lever.

One thing that strikes me about this is the lack of tactile feedback. Even Apple is working hard to fix that aspect of the buttonless phone, but with cars it's even more important. I don't really want to be fussing with a touchscreen while operating anything moving fast enough to kill someone (or cause $$$ in damages). Don't get me wrong I love touchscreens, but they aren't perfect (sweaty hands, gloves, etc...)

Totally agree. Looking at this car, I commend Tesla for going through with an insane idea like this, but personally I can't imagine having to deal with this every time I leave the house.

As noted in the article, the author speculates this is due to the fact that in the long run Tesla expects these vehicles to no longer require drivers. So the driver UI is transitional technology.

The idea that I would have to locate and press a graphical button to turn up the fan is pretty much a nightmare. I can do it without looking now. Even a quality voice-recognition interface would be lousy.

Not on a Model S you can't. Which is really annoying when you're in one as a driver or a passenger.

The steering wheel controls on Model S can control the fan.

I've noticed that passengers don't seem to have any problem using the touch screen.

If you watch DeMuro's review, have a look at the two 3D-wheely-buttons on the steering wheel. I think those would work well as a tactile interface - the only problem now is the mode selection. Couldn't a software update give you the mode selection on the left-right click on one of those things, and then you have 3 degrees of freedom to adjust with the remaining knob functions? Not saying this would be a perfect solution, as you still (1) have no good indicator for the mode other than the center touchscreen and (2) need two actions instead of one to do anything. On the other hand it would now be a complete hands-on-steering wheel based approach, so the tradeoff could be worth it.

When I test-drove the Model S, sitting there just made me ponder how texting and driving can be illegal, while giant iPad-like touch screens in cars are not.

Aren't they the same thing, in effect? I don't want to have to risk my life or anyone else's just so I can adjust the volume or temperature controls when I could have done that safely via a physical knob without averting my eyes.

Yeah, we need to stop seeing touch screens as "futuristic". They're a way to save money, replacing a purpose-built physical interface with a one-size-fits-all screen which can be manufactured at scale and programmed cheaply. Some things (e.g. cell phones) wouldn't be possible to make if they included purpose-built physical components for everything you might want to do with them, but in general, touch screens are worse than the interfaces they replace. This article on the subject is a good read: http://worrydream.com/ABriefRantOnTheFutureOfInteractionDesi...

During my internship on automotive cost engineering, I quickly learned that motorized anything on a car costs and weighs leagues more than its mechanical-only cousin. I am curious why Model 3 went down the path of total motorization, when it is marketed as a more affordable car.

Also, while I am becoming more open-minded towards touchscreen for non-realtime controls, E-brakes does NOT seem like one of them --- sure you don't need it that often, but when you do need it, you'd better pull it RIGHT NOW. Wiper is also questionable. If full autonomy works then sure I don't really need a clear view on the road, except it likely won't work as well when it pours and you need to blast the wiper to max RIGHT NOW.

Would it be more costly or less costly for Tesla to have just one set of underlying mechanisms for all of their cars at all luxury levels, or have different sets of mechanisms depending on luxury level? The motorized mechanism could be adapted to all levels of luxury by simply changing the controls.

>During my internship on automotive cost engineering, I quickly learned that motorized anything on a car costs and weighs leagues more than its mechanical-only cousin.

That is not true these days, eg electric windows are cheaper than manual, and electrically adjustable mirrors are cheaper than the mechanical remote mirrors.

As a principle, electric actuation enables the componentry to be less robust, since the forces are lower and more predictable. Also, you can standardise electrical actuation across several models and lower the cost, whereas different designs may be needed for the manual option (windows are a good example here).

Here's a reddit post [1] of a Model 3 owner locked out of his car because the 12v battery was not functioning properly. There is no physical key for the car. The Model 3 appears to be full of these questionable design decisions.

[1] https://www.reddit.com/r/teslamotors/comments/79rp5w/model_3...

A lot of cars are keyless.

https://en.wikipedia.org/wiki/Smart_key#Nomenclature

They [almost] all have a physical key. It typically slides out of the fob. I don't know of a single [non-Tesla] production car that lacks a manual way to enter. If my "keyless" Audi is completely and truly dead (batteries removed!), I can unlock it.

Update: changed to reference this is a Tesla rant, not specifically Model 3.

... Model S and Model X don't have keys, either. You're 5 years late making this complaint.

Sure, but that doesn't make it an invalid criticism of Tesla's design choices. Plus, every component in the Model 3 is far more centered around budgetary concerns and will simply be driven by far more people, so this problem is entirely likely to be worse on the 3 than previous vehicles. Any other car at this price point (or really most any other car) would have some physical key, and it's not automatically a reason to hail the car as visionary design.

I've gotten non-Tesla rental cars that didn't have a physical key, so there's that.

Are you positive? I believe there are other high end vehicles that don't have one, but in my experience even fobs that look like they can't possibly contain a physical key have one hidden inside somehow. Even if it can't start the car it can at least open a door.

What did they give you to start the car?

Lots of car fobs are just wireless transponders to authorize the remote start now, it's how those hacks to steal a BMW with a laptop work. But those fobs usually still have a real key inside.

My apologies, I should have said "Non-Tesla production car" as this was specifically a Tesla rant. The Model 3 goes to the next level though as there is not even a fob.

Well I don't have any way to get in my Prius if I lose my keyless fob, but maybe that is only one exception

So, can one of you Steve Jobs clone-prophets hurry up and advocate for some nice big rotary knobs and flip switches and try to get us over this 'no-physical' fetish?

Just wait a few years until everything is touch-screen'd, and then reveal it like it's a new idea.

"Wow! That knob has ten indents! One for every level of setting! I don't even need to look at the screen!"

Sell it as a UX feature, like everyone is selling this trend.

A 16 * 9 display is not even a suggestion of anything more than economy in action. That is the format flat panels are made in for an obvious reason.

Except that this isn't a standard 1920x1080. As the article mentions, it's slightly taller than that, leaving 1920x1080 after space is reserved along the bottom for the control strip. So it's a custom panel pretending a standard virtual height.

16:10 is probably the second-most common aspect ratio after 16:9, so not really what I'd consider "custom".

This is exactly the sort of broken design thinking that will make cars unsafe.

In one of the talks at Google I/O last year a VP from Audi (or Volvo?) spoke in a thick and lofty German accent about how "ve haf completely oferhauled ze driver exzperienze". He played a sexy video clip showcasing their new Android infotainment system (something like https://youtu.be/h_7_fKJ0PNs), and the first thing I noticed is how they'd taken away my traditional temperature knobs and replaced them with digital touchscreen ones. They looked just like physical ones, and were in the exact same place you would expect (https://9to5google.com/2017/05/15/android-cars-audi/).

So, I've gained absolutely nothing, and now I have to take my eyes off the road and look down at the stupid console just to change the temperature.

TLDR: Tacticle feedback is a Good Thing(tm) and designers should cultivate - not fight - muscle memory.

I don’t think Tesla did this simply to be edgy, dangerous or different. They did this to prepare people for a world without drivers.

This car was designed for the riding experience

One thing I've seen in certain very successful companies, is a pattern of proactively preparing the customers for the future. Apple, Amazon, and Tesla all exhibit this pattern. Microsoft has also tried to do this, though I'd say that Microsoft has erred on the side of annoying people too much while doing so. Come to think of it, Apple has erred on this side as well lately.

Most companies just react to customer demands. Certain companies actually have the foresight to lead what the customer will demand in the future.

I think this is a really interesting move from Tesla and it is making a good commentary on what the automotive industry has become. There's a lot of clutter in the modern car's UI, and I do agree that it could be slimmed down a lot. But the thing here is that Tesla is sacrificing usability for simplicity -- having to go through menus on the touch screen to access commonly used features like windshield wiper speed, relying on the screen in the middle to know whether your turn signal is on, or having to browse through menus to enable your parking break. You can't even see the speed your car is going on the dashboard, and instead have to move your focus to the thing that's basically an iPad mounted in the center console. The thing is, these are solved issues. Tesla could use a HUD to show even a simplified dashboard view. They could integrate the screen into the entire front seat/dashboard location/center console rather than just a rectangle in the middle. The only real answer to what they are doing is given in the article: they are trying to ditch the normal driving experience because they believe in the self-driving car as their end goal. This car is almost more of a commentary on this idea than a product (though I'm sure sales will be through the roof because Tesla has built itself as a trendy company the same way Apple did). The Model 3 is not for somebody who enjoys driving their car 24/7, and until self-driving cars become the norm it pretty much just exists to be edgy.

> You can't even see the speed your car is going on the dashboard, and instead have to move your focus to the thing that's basically an iPad mounted in the center console.

Center dashes have been a thing for a long time, my Saturn Ion had one and once I adjusted to it I preferred it since it was along the same axis as my side mirrors as opposed to having to look down.

this is a pretty weak analysis. His core notion that because the central UI is a 16:9 horizontal aspect ratio that differs from the model S, then it MUST be a passenger oriented design perspective as opposed to a driver specific one.

Well the same can be said that a horizontally oriented UI experience reduces the need for your FOV from looking further down and thus decrease time spend peering away from the windshield.

Design is so subjective. There are so many ways to interpret.. These articles proclaiming X instead of Y is hardly convincing at all.

> His core notion that because the central UI is a 16:9 horizontal aspect ratio that differs from the model S, then it MUST be a passenger oriented design perspective as opposed to a driver specific one.

Well, that and the fact the Model 3 gives the driver almost no controls.

I’m not sure if the model 3 has any voice-control, but a decent voice system might be a good way to augment the touch screen for some tasks like ‘wipers faster’, for example.

The voice control in the model is ok - it’s good at recognition, but limited in function.

They can at least enable this type of function later with software updates. This is a key differentiator with other auto makers, who rarely issue software updates after the car has shipped, and if they do, never include new features, and often charge you for things like map updates.

Reply on news.ycombinator.com