A Revolution Sweeping Railroads Upends How America Moves Its Stuff | Hacker News Reader

A Revolution Sweeping Railroads Upends How America Moves Its Stuff

morningstar.com

56 pointsycombonator34 comments

I do a lot of work with the rail industry in the UK (both heavy and light rail) - but the UK is much more passenger focussed.

Reducing cost can take a lot of tweaking to get right, and is likely to take more than one cycle to do. But new schedules do take some bedding in until they are running smoothly. Once that happens, you can then look for the next round of optimisations - tweaks to the way the trains are assembled in the yards which then start building the incremental gains.

If you have a better scheduling system, you build resilience into the schedule. You open up more time for maintenance, you keep your assets in better condition, so you have fewer breakdowns. One thing to remember is that (in general) there is a Pareto on train breakdowns - 20% of the locomotives cause 80% of the problems. It will be the worst locomotives that are taken out of service, so the reduction in service affecting failures should be quite significant.

You make some money on the OPEX side by doing more scheduled maintenance, and less emergency maintenance (overtime is really expensive!). You also make savings on the CAPEX side, because your assets last longer.

[edit]. Changed para order

Can you please explain why "new schedules do take some bedding" ?

Cant u sort of simulate the expected scenarios and get sort-of expected performance?

I realize there is unexpected events, but these will occur anyway in old and new schedules

Also - does real train software has nice graphics or do u get a boring excel ? ;(

In theory there is no difference between theory and practice...

As for software, the optimization of rail slot allocation is far beyond spreadhseets - for the whole French network it is a math-heavy process for which a whole team harnesses a bunch of specialist software. More about the process (in French): https://tel.archives-ouvertes.fr/tel-01423840/document

> In theory there is no difference between theory and practice...

This is really good.

Part of the problem in modeling is to also model the tens of thousands of humans interacting with the new schedule. Changing the train schedule also changes their schedules.

Hence, each time a schedule changes there is an uptick of accidents, disruptions, missed trains, overcrowding, etc until all adjust to this new timetable.

> I do a lot of work with the rail industry in the UK (both heavy and light rail) - but the UK is much more passenger focussed.

Same in France - allocating freight capacity in a passenger-focused network is tricky. A paper about that (https://tel.archives-ouvertes.fr/tel-01317522/file/TH2015PES...) - in French but with abstract in English:

"Train timetabling on a rail network is a complex task. In Europe, this issue lies within a process for which the infrastructure manager has to consider competing uses – passenger and freight trains as well as maintenance work – and several users of a scarce resource, infrastructure capacity. The aim of this research is to better understand the way this process is organised and works on the French rail network. We have chosen to tackle this issue focusing on freight as an activity which has been opened up to competition since 2006 but has experienced structural difficulties since the 1980s. The various time aspects of the capacity allocation process are the key focus of our analysis. Our approach considers the perspectives of the infrastructure manager and of their customers, foremost among which are train operating companies. We highlight that freight occupies a unique place in this process which is based on a schedule resulting from both European regulations and the legacy of decisions taken at the national level. The issue of the quality of the train paths offered by the infrastructure manager is studied with a focus on the links between short and long term requirements with regard to the train running deadline. In spite of their diversity, we demontsrate that freight operators need flexibility and visibility. Finally, we analyse catalogue paths as a specific offer that aims to guarantee capacity for freight. The German method, through the neXt project, provides an alternative insight into this issue. Accommodating freight train paths into the timetable is essentially a necessary but delicate balance of interests : standardisation vs. customisation."

Interestingly here on Long Island, NY we have the NY and Atlantic RR which is a freight operator who leases LIRR tracks. They move freight between mainland freight tracks in New York City all the way out to the "forks". They even revived some abandoned LIRR tracks in NYC and LI.

It's amusing to see a small freight train come rolling through a passenger station with platform. I don't know how they squeeze them in, probably only run off-peak.

This is very common in the UK which runs a mix (even if passenger trains do dominate.)

Our reluctance to spend the money to separate the passenger and freight traffic - even on the busiest lines - is part of the reason reliability of the network is often under question.

== is part of the reason reliability of the network is often under question. ==

The poor reliability is then used as a reason to limit future investment in passenger lines.

The customer perspective:[1] The metrics grain shippers use all got worse.

[1] http://www.grainnet.com/article/160328/precision-scheduled-r...

> CSX spokesman Bryan Tucker said the company could have done better communicating the changes to customers, but he defended the actions by pointing to its financial results.

I think the most telling part of the article

I don't understand: how/why is service worse? The article makes it clear that service is worse, but the details on why are important.

I can come up with a number of different scenarios, with different faults and restrictions. Without knowing deeper reasons I don't know if this is a bad idea (as BNSF claims) that needs to be scrapped or a great idea that just needs to get over growing pains.

This seems to be an argument that one scheduling system is worse than the previous. However, one of the major objectives of a scheduling system is to reduce cost, which this seems to have done. What is the downside?

You are missing the key pay of the question: reduce costs for whom? Since railroads are effectively a (regulated) monopoly reducing costs for Norfolk Southern or CSX can mean Wall Street profits combined with worse service for customers at the same price.

Wouldn't that be a question for regulators then in terms of how much profit can be extracted without lowering costs or using the savings on other investments?

I'm not sure I get the problem. Doesn't this just call for modeling a network of branches, nodes, with set of realistic max/min/stddev timings to traverse each, and an inventory + desired times of trains to be released into the system?

Is it that railroads don't have the expertise to do this? Hard to imagine.

Why are railroad execs huddled around a physical table + map trying to propose timings for trains to run through? Maybe the article is simplifying things that makes it sound really inefficient.

Is precise scheduling of trains really a revolution?

Perhaps not, but business practices fall out of style and get rediscovered over time.

This practice of leaving with or without your forecasted cargo reminds me of the software “release train” concept.

It sounds like something advanced manufacturers were doing in the late 1970s and that the laggards were talking about in the mid 1990s.

For cargo yes. Trucks arrival times are predictable within a few hours, trains are within days.

If trains are already running precise schedules on those tracks, then it's not that revolutionary really?

It's not a revolution per say, except that since it can often be hard for trains to bypass congestion/stoppages that schedules can go awry easily. In the US, there are huge bottlenecks around Chicago (where a crazy high percentage of all American rail freight passes through) and as much as the railways would love to build new track, various political/environmental issues have been preventing it.

As the article states, freight trains used to kind of run "on demand" centered around the customer, which made scheduling efficient train loads hard. Now they're trying to group into a more normal schedule, but this can still be hard as freight loads don't necessarily go point-to-point to their final destination and delays can cascade for switch-overs at hubs.

Source: Have been a long-time investor in a major railway and read the investor reports.

The article says:

> Now railroads want to run more like commercial airlines, where departure times are set. Factories, farms, mines or mills need to be ready or miss their trips.

You're saying: > In the US, there are huge bottlenecks around Chicago (where a crazy high percentage of all American rail freight passes through)

To me it sounds like they are already most of the way to the goal :D

So, basically a routing/allocation problem? You could throw SAT solvers at it, right?

The layoffs are going to hurt. Fewer good jobs for people with those skills.

Behind paywall unfortunately.

Try ‘web’ link.

What does this mean?

Most paywalled news sites have exceptions if you're coming from Google/FB. You can hit the "web" link under the title of the article on HN to Google search for it, and then click the article from there.

Has that ever worked for you with the WSJ paywall? Not for me.

One just worked for me for this specific article.

Yes, it worked for this article.

If you search for the article title you will find it on syndicate sites without the paywall, such as https://ih.advfn.com/stock-market/NYSE/norfolk-NSC/stock-new...

Perhaps a mod could update the link?

We changed the URL from https://www.wsj.com/articles/a-revolution-sweeping-railroads... to one that's publicly readable.

Reply on news.ycombinator.com