Legacy auto are almost all terrible at software. Tesla invented the softward defined vehicle (software controls almost everything, updates happen automatically over the air, software can add new features and they do it constantly, almost monthly). Ford, GM, VW, several Japanese brands like Toyota have all failed to make great software. With EVs more and more features (almost everything) is controlled by software. And legacy auto is terrible at it.
Tesla & Rivian have good ui, updates over the air. Tesla has a great 10 year history. Rivian had a good one year history until they made an extremely terrible update that made some people's cars undriveable. I don't recall Tesla ever having anything that bad. They lost their great reputation. Rivian needs at least a few years of continuing updates to recover their previously great rep.
Back to legacy auto. The failures of the blazer ev are apparent to anyone who uses one. Customers have been reporting problems with the blazer ev and other ultium cars for months on places like reddit group r/electricvehicles. The past week or 2 there have been a least a couple dozen reviews where gm sent out the cars to be evaluated in official reviews. And it doesn't seem like many or maybe any of them worked out well. This is a collosal failure of a huge auto company. The failures must have been apparent to everyone working there.
This is way too long already, I'm not going to go into vw software updates that take so long they were afraid it would run out the 12v battery system and brick the car (most all evs have a 12v battery that recharges from the big battery, but during os updates for safety it only runs on the 12v "starter"/utility battery).
If I can point to a single aspect which distinguishes the companies that are software first (google, apple, tesla, etc) and those that aren't: it's input lag on the UIs. Companies that treat software as an afterthought don't mind release software with horrendous input lag. First class software companies find that unacceptable.
Coincidentally I came to the same conclusion for video games; you can generally assume that a game won’t be much good if the menu options have input/animation lag — my theory being that any game designer who actually liked playing their game would have gotten fed up with it during playtesting
Having owned a pretty "old" tesla, a MCU 2 model S, the screen always felt responsive to be. All the MCU 1 cars were pretty laggy but they were also Tesla's very first screen attempt and I think they did pretty good for a car designed in 2009 or so
Sounds like something is interfering with the CAN bus and causing multiple corrupt packets from different modules.
That said, the article leaves a weird impression of being written by AI, or SEOed. Something about the repetitive sentence structure centered around the EV model name.
Electric cars seem like they should be so easy to make (they have less moving parts than ICE cars) yet almost all seem to have goofy problems like this. Why?
The current complaints about E-cars appear similar to those that surrounded ICE cars in the 1970's and 1980's. Such complaints were then mostly eliminated by Japanese manufacturers, leading to them replacing European and American manufacturers in many markets.
Perhaps the lesson is that absolute attention to the smallest details is needed to avoid producing 0.1%, or even 0.0001%, of vehicles that can cause severe reputational damage, i.e. an attitude quite different from the "move fast and break things" approach.
This has always been a problem for me. Why do they always need to design a completely new car, when they could basically just slot in the battery and motors?
Manufacturers could make a $20000 electric car that is equal to an equivalent-costing existing gas car (like a Camry or Corolla) in terms of capability. Reuse most of the same parts and technology from the original, just make it electric. Consumers would lap it up, not to mention how much it would help the environment. And yet, they don't, for seemingly no reason.
To be fair, there are differences between ICE cars and EVs. The battery and motors go under the body, not in front of it. But I'm sure they could do it if they wanted to.
All we get are premium-level cars that are slowly coming down in price, when what we really need (to get normal people to switch to electric) is a Camry-level electric car that is Camry-level cheap.
They want to build elaborate "subscription models" like some sort of web SaaS product.
"We do believe there are subscription revenue opportunities for us," Kummer said. GM Chief Executive Mary Barra is aiming for $20 billion to $25 billion in annual revenue from subscriptions by 2030."[0]
I can't overstate how unappealing this sounds, and how little faith I have in these companies to build "software on wheels" that works well, isn't buggy, doesn't have to be updated all the time, doesn't get in the way of the 1 thing I want the car to do: drive me where I want to go without interference.. while providing something necessary and functional that justifies subscription payments. Why would anyone want to pay Chevy for presumably what they can already get on their phones etc instead of just buying a normal functional car without pointless bugs?
>The battery and motors go under the body, not in front of it.
Doesn't that strike you as a rather large change? Significantly changing the center of mass on multiple axes and the allocation of space within the body of the car seems like a pretty big deal as far as the design of a car goes. And that's ignoring the fundamental differences in characteristics between an ICE and an electric motor and what effects that and the different placement has on the drive train and other systems.
Several manufacturers actually tried that but the results were garbage and the cars didn't sell well. It was mainly a shortcut to produce compliance vehicles for sale in jurisdictions such as California that required every manufacturer to sell a certain number of zero emission vehicles. The optimal packaging of batteries and electric motors is just too different to work well with vehicle architectures that were intended for ICE.
Even then, the Fiat 500e compliance car could have had a fairly large frunk had Fiat cared to not make giant plastic spacers to fill it out under the hood.
Compliance cars were just that, minimum viable products that manufacturers made plenty of crummy decisions on needlessly.
You’d think! I’m curious how this compares to the first few years of any entirely new model.
Perhaps it’s partly because it’s not just the same kinds of parts in different form factors and configurations. It’s an entirely new paradigm of parts for all the steps of the supply chain. A fraction of the parts, but still a brand new set of challenges.
Poorest quality parts, outsourced software development, outsourced hardware engineering, complete ambivalence on the part of the C-suite about the quality of their vehicles and future of the company.
They’re all brand new models. With ICE models, the drivetrains are rarely all-new. The platforms are shared across multiple models. So you rarely have a 100% new IVE car.
This feels quite alarmist. Could easily just be one relatively simple fault that’s generated a bunch of failure messages. Why not just wait and see before writing this big article about it?
This isn't some over-reaction. It's a colossal failure to execute from one of the leading car companies in the world. Their new platform evs are all having problems, this is just the latest one.
Their "new" ultium battery system is its own massive failure, they can't put batteries together for mass production, years after it was supposed to be working. Their previous technology platform worked and continues to work just fine, but they just discontinued the old bolt. New owners have been reporting problems for months. I figured they had improved problems before putting out a new vehicle, but this is just shockingly a complete disaster for them.
The Fully Charged Show has an excellent (and remarkably candid given his position) interview with Ford CEO Jim Farley explaining why legacy automakers, including Ford, struggle so much with software.
In short, they are trying to integrate modules written by their component makers, whereas Tesla (and presumably BYD and the other Chinese pure play EV companies) are writing all their software in house.
I think thats why Mercedes (and Tesla and Waymo) is eating the other legacy automakers guys’ lunch when it comes to self driving. Unlike the other automakers they have a huge in house division devoted to software in Seattle
Those employees are trapped in tens of hours of meetings every week with Mercedes Germany employees, and what code they do ship is NodeJS on Kubernetes on a crummy underpowered SoC to power the infotainment system.
There is a serious cultural mismatch between the Seattle office and Mercedes, just look at the Seattle offices commuting culture (mostly not by car, with many team members literally not owning cars).
It’s my launch ID.4 all over again. Wow. Genuinely did not anticipate that GM with the benefit of viewing that disaster would literally do the same thing years later.
We’ll never get it but I would love to look, in forensic detail, at the process that led to allowing the car out the door. Somewhere there is a deck or comparable vehicle where software was moved from red to yellow. I just cannot imagine taking this kind of brand and quality risk with such a key model. Massive organizational failure.
Low-level software engineers are all afraid to push the STOP button. They slightly softball their estimates, the impacts of their bugs, etc. Their supervisors soften it up just a bit more before reporting to their management. (This is all done via a byzantine tangle of Excel "project trackers" stored 7 folders deep on various Sharepoint sites.)
At the weekly meeting where the managers update on progress, they do a little bit more nerfing of the seriousness. They have faith the dates will be met, so all Criticals become Majors or even Minors. Majors become Minor or Warning.
Someone at the very top will get this shipped. They have no time to hear about failures, they want only Solutions. There are no Problems, only Opportunities. The wider employee base gets to attend monthly meetings with Powerpoints that show how "We're updating our Culture for a new Paradigm". What they take away from these meetings is to keep your head down and don't get in the way of The Paradigm Shift.
Only when the rubber meets the road (literally!) is all this exposed.
What gets me is how no one ever seems to see this coming. What is it about GM's[0] management structure that makes them so oblivious? There's got to be an arrangement of the incentive structure to avoid this all too common failure mode.
Legacy automakers are just big integrators and rarely develop their own core firmware. They just take the existing signals from all the different parts and program a conductor to get the symphony to play nicely together.
They will continue to get eaten by manufacturers that can control the entire stack, from servo firmware all the way up to infotainment.
Looks like it might just have a canbus fault somewhere. Maybe someone forgot a terminating resistor or curb the wire or crimped the terminal incorrectly. It also could just be poor error handling on the can network and being super sensitive to can faults.
My (limited) impression is that automakers treat software as a second class citizen in the automotive experience (so far).
At some point they're going to have to realize that the "Apple experience" matters for consumers and they should make the UX both usable and reliable.
At least one would hope this happens. Most likely one will, and then if it catches on with buyers as a purchasing decision then the others will hop on to try to remain competitive.
Tesla & Rivian have good ui, updates over the air. Tesla has a great 10 year history. Rivian had a good one year history until they made an extremely terrible update that made some people's cars undriveable. I don't recall Tesla ever having anything that bad. They lost their great reputation. Rivian needs at least a few years of continuing updates to recover their previously great rep.
Back to legacy auto. The failures of the blazer ev are apparent to anyone who uses one. Customers have been reporting problems with the blazer ev and other ultium cars for months on places like reddit group r/electricvehicles. The past week or 2 there have been a least a couple dozen reviews where gm sent out the cars to be evaluated in official reviews. And it doesn't seem like many or maybe any of them worked out well. This is a collosal failure of a huge auto company. The failures must have been apparent to everyone working there.
This is way too long already, I'm not going to go into vw software updates that take so long they were afraid it would run out the 12v battery system and brick the car (most all evs have a 12v battery that recharges from the big battery, but during os updates for safety it only runs on the 12v "starter"/utility battery).