I have strong doubt about that. I grew up in an automotive family and went to a heavily automotive school.

You basically have three major hold backs:

* Parts requiring manual assembly. My first summer job was at an auto supplier. It involved running 3 wires though a piece of conduit and wrapping it with electrical tape. It was a connector for some sedan brake light. Family friend. They still run the business the same way. Parts change too frequently and are too hard to automate.

* automation still requires maintenance. This isn’t just replacing parts on machines. This is also the entire supply of parts that go into fixing machines. Lines going down is incredibly expensive. Parts aren’t always readily available. There’s a whole industry of small “tool and die” shops that have the capability to build a part from specs next day.

* plants are huge and expensive., you don’t retool unless you have good justification. Typically, this happens with a model refresh.

These 3 bullet points are exactly correct and serve to support the claim in the comment I wrote which you are replying to.

For these reasons you highlight, auto plants in the US are generally as automated as is makes sense to do.

There's always room for more automation. Tesla, for example, is much more automated in many areas.

> Tesla, for example, is much more automated in many areas

By out-sourcing jobs to China.

> According to new figures released by the CPCA, Giga Shanghai was responsible for the production of 710,865 vehicles in 2022. And given Tesla sold 1,313,851 cars last year, that means 54.1 percent came from the Chinese plant

(I've worked on some automation projects for them)

They actually shoot themselves in the foot often.

They often over automate from the beginning, force too much, figure out it won't work, then just end up doing it how Ford does.

The closest I've seen to Tesla using more automation was just Musk spouting off random bullshit. Actual deep dives have shown that the automation tesla was trying to do isn't ready and they spent as long or longer fixing every car, often meaning they had lower output and more human involvement than competitors.

By what metric? The numbers I've seen indicates the opposite, although against European peers. But I don't know how reliable the numbers are, as the real data is probably a trade secret. The differences were really small however, and you might be able to slice them differently for a different result.

Their build quality issues and problems producing CyberTruck aren't exactly great advertisements for excessive automation.