It's fascinating to see the spread between different tram networks' terminal layover allocations. Layover times are a major driver of reliability, especially on higher-variability transit services. But they're also a product of infrastructure!

Fascinating finding. Railways around Tokyo rely on timed overtakes to provide exp and local service on 2 tk lines. These lines tend to have very high reliability (>95% <5m late), but there's apparently a *lot* of on-the-fly adjustment. More than 20% of overtakes do not take place where scheduled!

Visited Paris recently. A few transit thoughts: - The RER is one of the most incredible pieces of infrastructure...anywhere? - On-street trams are slow, even where the track is green and the urbanism is good - The region's efforts to design tram/metro/rail terminals for delay recovery _shows_

BART is planning changes for platform assignments at Daly City that will balance headways and enable more cross platform transfers at Bay Fair and MacArthur. bart.legistar.com/LegislationD...

As manufacturing continues moving southwards and outwards from cities, it's fascinating to watch new clusters emerge. A good example: the string of 5 steel mills north of Memphis. They're low-profile and (relatively) low-employment--but are now the most significant group of plants outside of Chicago

Today is Conrail day! 50 years on, I’d suggest we remember its least credited legacy: the Northeast Corridor. Without the region-wide network planning exercise that created Conrail, the NEC would still be full of freight—as it was until the 70s. Since Conrail, that feat has never been repeated.

This train is one of many that exports New York City's export trash to a network of landfills and incinerators across the eastern United States. As urban landfills reach capacity, arrangements like this one will likely become more common.

From my newest (and first!) substack post open.substack.com/pub/adenyaco...

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?

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...

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?

A few years ago, I wrote a long piece about why NS wanted to bulldoze part of a neighborhood on Chicago's South Side to expand an intermodal yard. homesignalblog.wordpress.com/2023/07/18/c...

Reading the UP/NS merger app alongside CP/KCS, you really get a sense for how railroad managements' read of the pax rail environment differs now vs 2021 In 2021, "expansion of psgr svc" made an appearance in the CEO's statement. Now? Just procedural assurances that the merger won't break anything

Such an elemental yet overlooked reality of planning. If you don't: - Design a new route to add value to the network overall - Make changes to the existing network to incorporate the new asset ...you're kneecapping yourself. A key part of transit hist: homesignalblog.wordpress.com/2024/07/12/t...

A few weeks ago, PATH announced a series of service expansions that will finally reverse this trend. Though paired with a rather steep fare increase, these changes will finally put weekend PATH service levels above where they were 25 years ago. Huge congrats to all the advicates who made this happen

This is insane. ICE raided an apartment building in South Shore (a 90%+ Black neighborhood) at 10pm, using chainsaws to gain access and throwing flash bangs. According to witnesses they just handcuffed everyone, with people who weren’t arrested being held for over five hours.

hello from the central valley

I created stringline diagrams for the Sonoma-Marin Area Rail Transit schedule, following up on discussions about why SMART can't run a 30 minute Takt. It's currently running a 32-minute Takt after opening the Novato Downtown and Petaluma North infill stations.

New story about trains and corporate mergers, and why you should never ever stop spending money on the railroads. medium.com/p/c6ab8375d4f0

Tangential to Stephen's point, but we really need to bring back the art of the high-profile government-sponsored market research report. At the height of Progressive scientific management fever dream-ism, this sort of thing would become a 1,000 page report called "The Grocery Question" or something

A lot has happened to PATH in the past 25 years -- 9/11, major construction work, COVID, etc. But consistent across all of these events have been ensuing reductions in off-pk service levels: since 2005, the number of PATH trips crossing the Hudson on Saturdays has fallen by *50%*

You really begin to see the divergence between the US and everywhere else when you compare this map (18% double track, 1950) to Italy, where about 26% of the network was double track in 1956.

At a high level, this is a good thing. Railroads suffer massively from the friction inherent to carrier handoffs. Mergers can mitigate that, potentially opening new markets and allowing more effective competition w already continent-spanning trucking companies.

A few months ago, I paid a visit to Boston and spent some quality time with the MBTA’s bus service. After spending a while digging through T’s (excellent) open data, I had some Thoughts to share about the experience—and bus operations in general. homesignalblog.wordpress.com/2025/06/29/h...

Going to publish a much longer look at bus operations soon, but it is just _incredible_ how significant the trip-level effects on performance are, even on fairly frequent routes. Two examples from the MBTA's network:

For my money, one of the greatest gaps in benchmarking practice surrounds the *tactics* of service management — how different systems respond to incidents, gaps, congestion and so on. Gaining that knowledge just got a good bit easier, thanks to this incredible new tool:

No only there will be zero accountability. There will be zero lessons learned from the errors of this project because we will never know what went wrong. So we will start from zero and make the same mistakes again and again and again...

Continuing my series of posts on the railroad industry's immense challenges managing complexity and market alignment, I have (finally) finished some thoughts on the trajectory of railroading's growth engine: intermodal homesignalblog.wordpress.com/2025/05/11/w...

“Mommy, no. Mommy.” Daughter captures the moment masked ICE agents smash the window of her mother's car in order to take her into custody. The government claims the 52-year old seamstress is an associate of the MS-13 gang. www.thebaltimorebanner.com/community/lo...

This is an excellent post, well worth your time to read. When contrasting this history to the US, it's hard to overstate the deleterious impacts of fragmentation on investment. Marco's emphasis on the importance of nationalization is critical!

Theoretically, dispatching buses to a timetable w/ accurate running times should yield similar levels of service consistency on most routes as pure headway management (so long as timepoints are well-placed). But I feel like I'm missing something--anyone have reading recs?

Consider this. 6 people in plainclothes approach you, and begin to cover their faces. At least one appears armed. As they grab your phone and backpack, they do not prove who they are. They say they're going to take you, but they're not saying why. What would you call that?

read the quote. then read the date at top. then ponder the nature of railroad management and investment markets.

The whole case for the expansion is predicated on two incorrect suppositions about traffic trends - Truck traffic growing due to port expansion [it hasn't] - Car volumes growing due to development [they're flat vs 2010, when JC/Bayonne were 15% smaller] homesignalblog.wordpress.com/2023/07/08/d...

Here is my latest long post on Substack about the Long Modernization of the Italian Railways. I tried to capture how the railways in Italy (and elsewhere) remained relevant to today's world and what this history teaches us. I hope you'll enjoy it. open.substack.com/pub/marcochi...

siri, what’s a strawman? www.aar.org/news/study-c...

It's a lost nuance in transit's financial history, but these observations on the relationship between off-peak demand and overall profitability remain extremely relevant to system design discussions today:

well this is quite the anecdote

scenes from the chemical coast

These constant fights over passenger priority are depressing on several levels. Of course, freight railroad mistreatment of passenger svc is bad--but the convo currently misses that it's essentially inevitable with our current rail operations paradigm in the US. www.trains.com/trn/news-rev...

"Commuting from downtown Washington DC to jobs in adjacent Montgomery County, in 1968, took between 43 and 72 minutes, depending upon the destination and assuming minimal waiting time...A suburban resident using the same route in the opposite direction spent 12 to 27 minutes less in transit."

In 2004, when Jersey City was ~20% smaller: - PATH ran the normal weekday service pattern (at 15 min frequencies) on weekends, plus 10 min midday service on all lines. - HBLR ran 15-minute headways middays & weekends (vs 20 now)

if i had a nickel for each time i've visited intrepid's potash mine near moab, i'd have two nickels

In Chicago: “The planned Red Line Extension is projected to cost seven times more per mile than the Orange Line, which we completed in 1993 (that’s after adjusting for inflation).”

One map I have grown to hate: this one. Nostalgia for all the lost intercity trains often masks the fact that those services were dinosaurs. Some stats: - In 1954, 70% of e/b trains on the PRR from PGH left PGH 10p-5a - In 1960, 50% of the Santa Fe's Chicago-KC trains arrived in KC 10p-5a

not to beat a dead horse, but: it’s consistent across 100+ years of railroads (re-)“discovering” the benefits of network planning that these plans are compromised by incomplete scheduling, imprecision, and a belief that vulgar cost minimization is OK homesignalblog.wordpress.com/2021/12/27/e...

There’s an interesting analogy to be made btwn transit and freight rail networks. Historically, several transit agencies have (over-)emphasized suburb-to-downtown service over a more comprehensive route network because it provides strong loadings and isn’t that hard to run well

allow me to put in a good word for the california zephyr