We’re used to dashboards, to keeping an eye on resource-levels as indicators of the overall health of a system.
We’re less used to understanding or tracking the processes that update those levels.
If we did, we could probably fix things more quickly when they go wrong.
If we made those processes self-managing (as they often are in modern cars), not necessarily automated, we might even prevent things going wrong in the first place. After all, humans are sensitive instruments.