If you're curious about learning more about PSR as it was intended to be implemented, I can't recommend Justin Roczniak's writing (and podcasting) around it. He wrote essentially the same article for the New York Times not too long ago[0], and his longer-form podcast discussion is even better if you have the time[1].

In his telling (and I'm paraphrasing heavily here), PSR — as it was originally designed — was meant to get everything on precise scheduling (surprise), thereby allowing you to better predict when and where you'd need workers well in advance. That would let you better optimize not just your trains but also your workforce. If you have a schedule, you can make optimizations that keep train cars from sitting around, and your workers reap the benefits of a fixed schedule (time off, less on-call time, things like that).

In practice, getting things to run on a schedule often meant capital costs that a of these companies balked at, like upgrades to rail yards and track maintenance to keep things running on schedule. The only real part of P.S.R. that ended up getting adopted, as this article points out, is the optimization part. But without the capital investments, you don't get a precise schedule, nor do you get any of its implied benefits.

Optimizations ended up meaning things like pulling more cars with fewer trains. I don't think the article mentions this in explicit terms, but these longer trains are often a few miles long. This length means the trains don't fit in rail yards, nor do they fit in the sidings that allow Amtrak trains to pass them (which in turn keeps Amtrak trains from running on time). Since trains don't fit in the yards, they have to leave as soon as possible, and since there's no fixed scheduling, "as soon as possible" ends up meaning any time at all with little notice, meaning engineers are on call almost all the time. Plus, when you have a train that's almost 3 miles long, they're more likely to derail on curves, and staff have to walk literal miles to fix problems along the train.

There's a whole lot more to this than I can repeat off-hand on a Friday afternoon, but this issue isn't going to go away. The rail industry is untenable, and given how scared everyone is of a rail strike, it also isn't going away any time soon.

[0]: https://www.nytimes.com/2022/10/09/opinion/business-economic... [1]:https://www.youtube.com/watch?v=69A_UCdikE8