Pedestrian Observations

Right now it takes between 100 and 120 minutes (scheduled) to go from Poughkeepsie to Grand Central Station or vice versa during peak times. For Riverhead to/from Penn that's 120 - 130. Add travel time on both ends to get door to door times, plus variability on the high side (very little chance of being early).

Most people consider that uncommutable though people do it.

Poughkeepsie is 73 track miles from GCS and Riverhead is 75.1 track miles to Penn Station. The average speeds that would be need to bring these areas into the reasonable commuting range are not Shinkansen-esque.

Even within the official borders of NYC there are some areas with quite difficult commutes to Manhattan. Consider the Bay Terrace neighborhood of Queens. From there to the NYSE stock exchange via public transportation google maps recommends the Q13 bus to the Bayside LIRR station, a Port Washington branch train to Penn Station, and then the number 2 subway. Total trip length: 130 minutes, monthly cost: $330.14. Or consider co-op city in the bronx, from there to the stock exchange google maps recommends a .9 mile walk to the 5 train and then an hour long ride on that subway. If walking that distance was a problem you could take the Bx30 and then the 2 subway for a total trip length of 106 minutes.

There are many many sensible transit projects. The big problem is that we have seemingly lost the ability to do transit projects in any sort of reasonable cost or time in the greater NYC area. I've posted…

Piecing together the story from https://pedestrianobservations.com and other places, I would not imitate Japan in this regard.

- Japan's terrain is fairly ideal for rail. Yes, mountains make for more work, put people are also naturally congregated near the coast, in valleys, etc. Additionally, the largest cities with the high population form a linear chain from roughly Nagasaki to Tokyo. That means more network effects per distance of rail (there's no penalty separate than time delay for traveling through other cities first).

- Japan has perhaps the best policy around land use in the developed world. See https://www.sightline.org/2021/03/25/yes-other-countries-do-housing-better-case-1-japan/ for some details. This helps greatly with density and clustering around rail stations.

- The privatization was triggered by the fright around Japan's big deficits after the 80s boom, and perhaps still maintained with some fear of that. That doesn't make it a good idea, of course, but I'm willing to call it more "defensive neoliberalism" than "offensive neoliberalism", but maybe that makes it a little less ideological given that Japan was the poster child for all the neoclassical economic doomsday predictions?

- https://pedestrianobservations.com in particular says the privatization and and general being spooked about more debt holds back some construction on paper, but since the rail is so well established already and operationally profitable, there is less to screw up.

In…

I have been hacking on a better calculation environment for train speed zones.
Alon Levy wrote some python code that calculates how long it takes a train to accelerate and decelerate between different speed zones.

I have been cleaning up the code and adding more to it. I'm bringing it into a jupyter notebook so transit advocates can better present their routings. So far I have a framework that allows easy expression of alternate options along a segment of track

I'm not far off from being able to click and drag track segments on top of a map with ipyleaflet. I also want to hook in an optimizer to recommend the best track routings by cost.

There's still a lot of work to do. Is anyone else interested in hacking on this stuff?

[1] https://pedestrianobservations.com/2023/10/27/setting-speed-zones/
[2] https://github.com/paddymul/train-calculator/blob/main/Segment-generator.ipynb

"why is American infrastructure so bad?" has been a topic I am curious about. The above article does not answer the question.

I'd prefer not to care about American infrastructure costs. Unfortunately, I occasionally have to leave my house, and the consequences of bad infrastructure stare me in the face: people living in their cars, traffic jams.

The best attempt to actually answer the question I've seen has been "Pedestrian Observations" blog. Their take on the same GAO report is:

[New Report on Construction Costs Misses the Mark](https://pedestrianobservations.com/2019/07/22/new-report-on-construction-costs-misses-the-mark/)

"The GAO report is out now, and unfortunately it is a total miss, for essentially the same reason the RPA’s report was a miss: it did not go outside the American (and to some extent rest-of-Anglosphere) comfort zone"

As a counterpoint, I'd recommend reading Alon Levy's blog post "Why the Focus On Penn Station?"[1]. The TL;DR is that the 650,000/day ridership number significantly overstates actual ridership, and includes subway ridership that wouldn't really benefit from most proposed improvements.

There are far better uses for transit funding in the NYC area. Even within Penn Station, the low-hanging fruit would come by improving the track level (wider platforms/fewer tracks), rather than the concourses.

If this piques your interest, Levy's "Eliminate Penn Station" proposal[2] is worth a read, although he admits it's somewhat trollish.

[1] https://pedestrianobservations.wordpress.com/2013/05/31/quick-note-why-the-focus-on-penn-station/
[2] https://pedestrianobservations.wordpress.com/2015/07/15/eliminate-penn-station/

What's interesting is dozens of message on HackerNews and nobody links to any type of studies or evaluation of Elon's work, just repeating is selling point that because the tunnel is smaller, it will be cheaper. Public transportation and cost is a very well studied topic internationally (outside the US) and we can't just assume that for a fact. Actually, boring is actually not the most expensive part of building a subway. Even this idea of saving 10x the amount of money, actually just works when you compare with other US subway construction projects. Close to 10x saving has already been achieved in Spain and in South Korea, and surprise, it didn't need some new fancy technology.

Great article here: https://pedestrianobservations.com/2017/12/15/elon-musks-ideas-about-transportation-are-boring/
Basically