I work for another EV manufacturer. Good God I would love to switch from CAN to ethernet. We already have way too many CAN buses and we're considering adding more. But it will take decades for our suppliers to even consider the switch. We just paid Bosch $15M for an upgrade to our brakes, they promised the change in six months, it took them two years and quite a few additional millions of dollars, and it's filled with bugs. We can't even run the new brake firmware on public roads it has so many bugs. And we'll have to pay them millions more and wait months or years more for them to fix their own mistakes. This is why Tesla does almost everything in house. I wish my company had the resources to.
Well, then you should maybe ask why do you need such complexity, when cars 5-10 years did not needed to do exactly same task - taking a person from point A to point B.
First of all, there are regulatory issues, that continually demand more both in terms of efficiency and in terms of safety. System that were innovative 10 years ago are not a requirements and new system are recommended and will soon be required. And that is just for minimal compliance. If you actually want to be up to date on safety systems you need even more.
Second, if you are a company you need to actually sell product. And turns out costumers don't want technology from 10-20 years ago. Costumers actually buy stuff with more technology in it. No matter if people on HN rather drive a Honda from early 90s.
So to just tell a company 'just go back in time and that will solve your problem' is simply not gone convince anybody.
Break-by-wire is pretty standard by now. Unless you have any evidence that the system Tesla uses is worse then traditional automotive CAM and that it is so much worse that there will be lots of Cybertrucks unable to break then I'm just gone go with team of engineers who designed a new system after having decades of experience with the older system rather then some guy on HN.
