The $25k car is going extinct?

I work on my own cars now (as a hobby really) and one of the reasons the new cars are so expensive is they are much more complicated. A lot of this seems to be over-engineering IMO. This is alluded to in the article, but not explicitly stated.

The cars I work on are from the early 90s and everything is very simple to understand.

e.g. Electronics are normally simple circuits that aren't much more complicated than what you would find in a door bell and finding faults is normally just tracing wires and using a multi-meter. I had issues with the brake lights / reverse lights not working, the issue turned out that the spade like connector in the fuse box was pushed through and was making partial contact. Price to fix this was £0.

EDIT: Just remembered this isn't accurate. I had to buy a new reverse light. The entire reverse light assembly was ~£20. So the price to fix was about £20. The light assembly itself was like a big bicycle light.

My newer car needs a OB-II scanner to diagnose anything with a phone app. While this is arguably quicker it can be misleading. Sometimes it will be telling you that something is malfunctioning but it is really the sensor itself. These sensors are £200-£300 a piece. Replacing 4 glow plug sensors cost me £800. I was paying essentially to make the "you must service your engine" light to go away. There was nothing wrong with engine itself.

I think in time we see a similar trend with EVs. They are, by many metrics, vastly less complicated in terms of hardware. Software, of course, is another matter.

Software doesn't have to be complex. The most complex piece of software EVs absolutely need is the BMS and the charging protocol.

Everything else is super basic. There's a reason some of the earliest vehicles were EVs.

You think that software isn't complex because you've never seen it.

What should it do when the throttle pedal goes from 0 to 99 percent? That's likely an electrical issue, not a driver command to plow through the school zone. I could probably think of a dozen such scenarios, and the true number is probably in the hundreds. They all have to be proofed mathematically. With redundancy.

Out of eye, out of mind.

There is fundamentally no difference between a golf car and an electric car other than a higher top speed and more powerful motor.

> What should it do when the throttle pedal goes from 0 to 99 percent?

How fast? 0->99% in 1 second is likely the user gunning it. 0 to 99 percent in a millisecond is likely a fault. In either case, the simplest solution is a capacitor in-between the signal and throttle. Doesn't need to be particularly beefy to get the job done.

The problem is you are thinking about this as a software problem when it's an electrical problem. There are a lot of electrical components that have instantaneous response times, well known curves, and perform exactly the jobs you'd want faster than what you can do with software.

You want to minimize the amount of software between the accelerator and the motor precisely because you want to make the car as responsive as possible. Putting software in the middle creates delays and needs for very complex real time software and more expensive components.

But an EV has instant torque; going 0-99% in one second is probably unwise and not fixed by a capacitor. Software's what helps us not strip the rubber off the tires, or mitigates a slipping wheel on ice. It's a lot more than a capacitor at work.

This is a belt-and-suspenders thing.

The gas pedal on an EV isn't connected to a passive rheostat gating the entire power output of the vehicle.

It's a low-voltage sensor. A capacitor can swallow a transient of a 0.1v signal bursting to 5v for ten milliseconds, and then software converts it into whatever "sport mode acceleration curve" the marketing department calls for.