Uday Schultz

and they say it isn’t real life 😤

It's going to be awfully funny when UPNS's flagship trains run into the sort of operational split incentive problems that *Vena himself* cited as a reason we needed mergers in the first place

www.trains.com/pro/freight/...

Yup this!

I wouldn't want to purport to have any especially keen insight into LA Metro's performance processes, but: you'd be surprised at how incomplete feedback loops can be in these sorts of orgs! Both institutionally, and analytically.

But for an empiric example, see here. First train did great time, but the second did not, because it got delayed waiting for the first train to clear. To get 4-4.5 mins in both directions, you need trains to be free running...which in turn requires more margin in the headway

Correct, yes. As for the math...I wish I knew enough about the specifics of the signal system out there to say! This is where issues like block length and the nitty gritty of interlocking logic and all become super important.

I'll readily admit, though, that my estimate could have been more precise. For file size reasons I requested stop-level data from @transsee.bsky.social, but then had to rely on stringline analysis to get the paramount-norwalk times.

that, plus the following:
- it takes non-zero amts of time to move switches
- diverging moves are slower than straight
- depending on where circuit boundaries are, trains may not release switches as soon as they're off of them

...is how you get to 6 mins

switch-switch runtime isn't really what you're planning around. if you want this to run cleanly, what you want is for a train approaching the single track area to get a clean route through to the other side without slowing for conflicts, which effectively means the single track starts further back

Great question! To my knowledge, the outage was interlocking-interlocking, with both pockets at Norwalk avail. When I looked at Youtube videos of the outage, it indeed seemed that switch-switch runtimes were around 4.5 mins.

*however*...

Yes, and close/deprioritize Memphis IINM

I wish. Service in that lane has basically always been good, to the point that *pre-merger* NS could advertise faster cutoff/avail times than you're seeing there.

To be fair to NS, the total scope of the investment was significantly less than they’d hoped for.

But to be fair to the states, even with that limited $ the service was once quite effective! They had travel times down to 35ish hours iinm

There may have also been some finding put towards line capacity improvements in VA and TN, don’t remember off the top of my head.

I believe most of the govt money went to new/expanded intermodal terminals at Greencastle (since closed), Charlotte, Memphis and Harrisburg. NS also funded a new terminal in Birmingham, which they’ve also since closed.

(first quote: www.proquest.com/magazines/la...

second screencap: www.norfolksouthern.com/en/ship-by-r...)

I may write about this more formally at some point, but the degree to which NS's cut-service-and-hike-rates strategy completely kneecapped their Crescent Corridor partnership with states to decongest I-81 is nuts. After billions in govt investment, train service is now *slower* than it was before.

So, Brendan, Why is the Millennium Line Bad?

A Study in Schedule Design From time to time, I enjoy browsing the LA Metro subreddit. A creature of the East Coast, I am forever fascinated by the Los Angeles’s preculiar mix of ambition and ambivalence around transit…

Earlier this fall, I noticed something odd going on with rail service on Los Angeles's C line. So, I wrote about it, because I could not have asked for a more perfect encapsulation of how schedule design can drive service outcomes.

homesignalblog.wordpress.com/2026/01/03/a...

Thanks for the detail here! Makes a ton of sense.

Gonna disagree with you re the problem though — this specific situation is temporary, but basic issues planning non-revenue movements are gonna affect a lot more than just this

I will grant that surface transit ops are legitimately difficult, and fixes do often require support from a level of decision-makers above agency ops folks. But the existence of such constraints makes it *even more important* that agencies manage the controllable...

a stringline chart showing a bunch of streetcars entering service behind schedule

All the Toronto streetcar discourse has had me looking at a lot of TTC stringlines, and I must say, it seems there's an incredibly opportunity for quick improvement there with some basic terminal ops fixes.

Why was nearly every streetcar *entering service* on this random wednesday late?

On a related note, I'll be curious to see whether any of the states take NS to task for the Crescent Corridor. All that money spent, and then NS:
- Closes the terminals built in Greencastle and Birmingham
- Transforms their NE-Memphis train from a priority svc to a plodding milk run

I'm excited by UP's proposal to shift traffic onto the Wabash, of course, but Chicago flows aren't going away and the use of the Lincoln Svc corridor for said flows has acquired a certain degree of path dependence thru the construction of Global 4

...we'd be in much better shape for passenger service on that corridor, the St. Louis terminal would be (marginally) less congested, and the BNSF transcon would have a stronger competitor. Alas.

Had we either:
- Negotiated an access agreement for UP to reroute more of their CA-Chicago traffic away from this line
- Exercised a little foresight in the '80s when considering whether or not to abandon the eastern link of UP's route from NM to KC
x.com/A320Lga/stat...

This one was always such a failure of freight planning. SP bought the route from a regional because they were desparate for their own way into Chicago. But passing through St. Louis sucks, so UP's priority trains go via BNSF already.

it is! actually pretty easy to query out. but yes, the hard part would be then joining those data to rail lines...

it's not going to be perfectly accurate, but it provides a fairly decent directional sense of what's up. just would be so much GIS work.

If you want to spend several months doing data cleaning, it is technically possible to make one for US railroads using the FRA grade crossing data