@CoolSWEng I’m inclined to believe this is almost certainly not the reason.
For one thing, the odds of having a design team with no drivers in it are just pretty low. I work in tech, and in every design team I’ve ever been in I was the only non-driver. Most of the world is still too car-centric.
For a second thing, lots of good designers very often design for use cases which they themselves don’t share. You don’t have to be a pilot to design an aeroplane cockpit.
To me it sounds crushingly plausible that a car company could end up with a tech team that has few enough drivers in it that they could be overruled by the others, plus a management team that "leads by spreadsheet" enough to not know enough to overrule the tech team's design.
@rastilin @CoolSWEng Most of the designers I have worked with were not the target users of their designs. And if by chance they were, we took steps to remind everyone that “you are not your user”. It’s neither necessary nor desirable for good design work to have a design team of ‘insiders’.
I don’t believe that the answer to bad ergonomics in modern cars is “hire only designers who are petrol heads”.
@maccruiskeen Which is a fair criticism, but good designers should take that kind of thing into account whether they themselves are a publisher or not.
I’ve also used systems created by designers and non-designers that are very poorly optimised for any use case but the one that was most pressing at the time it was made.
There are all sorts of reasons that bad designs come to be though.
@paddyduke @CoolSWEng @maccruiskeen
Is it really not desirable for the designer to understand their users and product? I feel this entire thread is effectively about the consequences of designers that don't use their own product.
I feel that it often boils down to the question of "if you yourself wouldn't use your own product, why should anyone else"?
@rastilin That’s a straw man. The designer should understand the use cases of anything they design. But that’s not remotely the same thing as “being the person you’re designing for”.
If I designed things based on whether I would use it, I would probably design bad products from the perspective of everyone else. And I would absolutely bet that the designers who created these terrible car interfaces believed that they themselves would use them.
It's an appeal to authority to say that because the theory says it's good, then it must be good. Presumably the users who hate it aren't experienced enough in design to understand why the interfaces don't suck. Also if those are the designers that don't drive and have little experience with cars, does it matter what they believe they would use?
Though I suspect that yes, the fact that touchscreens can be cheaper than buttons probably also factored into the decision.Though even then, surely they can't be *that* cheap, plastic buttons are basically free.
@maccruiskeen 100%
Most likely it’s a multitude of factors encompassing the designers themselves, the environment and resources available to them, and pressures from other parts of the business.
I’m only saying that I don’t believe the whole problem is down to a handful of user interface designers not being enthusiastic drivers.
The touchscreen was going to be in the car regardless for other UX reasons. It also can perform the functions of the physical buttons. The cheap/free buttons still need to be manufactured, stored and fitted. Plus, all those buttons need electronics which also has the above costs.
If you look at it purely on a spreadsheet it makes immediate sense.
> You don’t have to be a pilot to design an aeroplane cockpit.
Yes, you do. In all major airplane companies user interface design decisions directly involve the end users (i.e. pilots); it's a lesson learnt the hard and bloody way.
@datenwolf @CoolSWEng You need a pilot (preferably multiple pilots) involved. But the designer usually works as a designer and is not necessarily a pilot themselves. I’m in no way arguing that you shouldn’t get input from your users. That would be absurd.
If you had to hold a role in order to design for it there would be no designers or architects or planners. And yet here we are. Design is a specific skill set and just having domain expertise doesn’t necessarily make you a good designer.
@paddyduke @datenwolf @CoolSWEng
"The worst user interfaces are created by software developers. Except those created by the customer."*
*And especially customer management.
@paddyduke @CoolSWEng Yeah. It seems dubious to blame all the individual designers and not the leadership and the structures involved.
Even if the designers aren't the target market for the product, are they revising designs based on feedback from people testing the product? Who decided that essential functionality of the car would be on a touch screen in the first place? If the CEO and other top executives weren't involved in deciding that, they are nevertheless still responsible for recognizing if this might be a problem. (What are they paid so much for otherwise?)
If you just ask a team of designers to design a touch screen interface for these things, then just implement it, without any particularly testing and iteration or any thought about if it's a good idea in the first place... It's not really the designers who caused the problem here.
@CoolSWEng
Okay, so they KNOW, that it is a problem to produce a "2-tonne projectile in a world of [...] targets". Great.
Agreed.
The US Navy will replace the touchscreen throttle and helm controls currently installed in its destroyers with mechanical ones, says USNI News. The move comes after the release of an accident report from the National Transportation Safety Board about a collision between the USS John S. McCain and a tanker ship in 2017, which cited the controls as a factor in the accident.
@CoolSWEng When we were “shopping around” for our first EV in 2022, I briefly considered, and test-drove, a Tesla. (The words “bullet” and “dodged” spring to mind…)
The main reason I HATED the experience: there were hardly any “hardware” controls, and so much was done on a large touchscreen. There are reasons why car interface design hasn’t changed much in decades… muscle-memory is vitally important when you’re a ton of metal barrelling down a road 🤦🏻♂️
@CoolSWEng I think the primary reason for modern cars lacking physical controls is that the billionaire owners of the car companies have chauffers instead of driving themselves.
Those stock holders then hire the CEO that promises to save the stock owners money by removing all physical car controls and replacing them with a touch screen. Those stock owners who make the decision which CEO to hire, won't know that removing the physical controls is a stupid idea because they don't drive themselves
@harmone oooh, I hadn't seen the parallels between the demise of Word Perfect / rise of Word and the demise of physical controls in automobiles before, but now it's staring me right in the face. Hard.
@CoolSWEng Tesla originally used a big touchscreen because they couldn’t afford to have tools and dies made for injection molding of all these controls. Then everyone else cargo-culled them because it looks cool.
In Europe at least, cars won’t get a five star rating if they don’t have sufficient physical controls:
This seems to miss some context. The UX engineer are not deliberating wether a touchscreen is a good idea instead of a button, they are focussed on the UX on the touchscreen. I work in a tightly regulated field where there are major consequences when things go south - decisions like these are not made by engineers, who are trained to look through the potential pitfalls.
The decision to use a touchscreen - and that particular touchscreen (with whatever specs it has) - is something that is not up to the UX engineer. And, even if it did come from the UX engineers, there's a whole level of middle-managenent that has had to sign-off on it.
This decision is simply cost based: one part that you were going to have in the car in any case, that can just do all the functions that all those buttons would do ... the savings are easily visible.
@CoolSWEng
Is it just the touchscreens, or more generally the software and who owns/runs it, which allows 100% surveillance, free training of A"I"s, etc.?
How far are we from seeing a "car as a service" money making scheme, where the owners won't own the car but instead pay and get a license by the company to merely use it (like software)?
@65dBnoise @CoolSWEng Former tech exec car industry here:
1. It’s a combination of the infotainment and vehicle automation systems, including over-the-air software updates for both.
2. Not far, but the traditional automakers are still very hardware oriented. They want to sell you a tangible asset. They’re not good at services in my experience. Your idea is ripe for exploitation, especially when people can come together in small groups to cover the costs.
@CoolSWEng With respect, I think this comment says half of it. You don't push a car out just like that, designers don't just put their tablet in the car and nobody ever checks it again. People all through the chain didn't care.
And if nobody ever checked, that's also the other half.
@CoolSWEng Whilst I like the simple explanation, I think this is not the root cause.
I think the entire industry is trying so sell that "abstraction from the broader reality" and for some reason the vast majority of us want to buy the dream.
So this is embeded in not only the UI but also the external design (SUV that make it impossible for drivers to see children), the advertising, the reviews, the infrastructure.
@CoolSWEng I'm an engineer and sewer. Reformed (?) Collector of sewing machines.
I've got a high-tech, modern, $7k sewing and embroidery machine that is just an awful experience. There is a plastic seam that the thread will get caught on on the way to the thread snip. It's got a wide deck for things like quilts, but the spaces where your fingers are needed just a bit too small. It just isn't comfortable to sew with.
None of my vintage machines are that poorly designed. There is more clearance around things that need threading. They make precise sewing easier, because I have analog controls instead of pre-determined computer controlled increments. A thread width over matters, and is visible in the results.
My daily sewer is a 1950 Featherweight that does 2 things exceptionally well: sew backward and forward. I rarely use my experensive machine.
The basics get lost in the noise of features, if you ask me. Which is my my car will give you the Gs on a corner but lacks a max defrost setting.
@CoolSWEng The ultimate in ergonomics design used to be considered to be the motorbike where, by necessity, you have to be able to work every control with a small movement of a hand or finger or foot without looking at them.
Is that still the case?
I want to share but the terms and conditions at the top of the alt text seem pretty disrespectful to the concept of alt text. I want to blame FT for that
@CoolSWEng That matches with what I've heard some years ago:
No driver's license, no interest in cars but implementing the code for the "turn indicator" on left side of the steering wheel, first SW version:
- "turn indicator up": lights outside flash one after the other "clock wise"
- "turn indicator down": lights outside flash one after the other "counter-clock wise"
@CoolSWEng Finally! The first time I saw a panel in a car, I thought: That's going to cause accidents.
I rented a car where the A/C controls were _sliders_ on a screen. And you *had* to stare at it (and not the road), just to adjust it. I finally had to pull over and stop the car just to set the temp. And as I drove further, and night fell, I had to keep pulling over to re-adjust.
I will never buy such a car.