How Learning from Japan Could Transform Our Railways
HTML-код
- Опубликовано: 11 июн 2024
- Watch this video ad-free on Nebula: nebula.tv/videos/rmtransit-it...
Often times, we might think that a flexible railway network especially at a large city centre station will allow for better network management, but countries with far better railways than us like Japan is proving why simplicity is better.
Read the very interesting research paper comparing Japan & the Netherlands' railways: www.cvs-congres.nl/cvspdfdocs...
As always, leave a comment down below if you have ideas for our future videos. Like, subscribe, and hit the bell icon so you won't miss my next video!
=PATREON=
If you'd like to help me make more videos & get exclusive behind the scenes access and early video releases, consider supporting my Patreon! Every dollar goes towards helping my channel grow & reach more people.
Patreon: / rmtransit
=ATTRIBUTION=
Epidemic Sound (Affiliate Link): share.epidemicsound.com/nptgfg
Nexa from Fontfabric.com
Map Data © OpenStreetMap contributors: www.openstreetmap.org/copyright
=COMMUNITY DISCORD SERVER=
Discord Server: / discord
(Not officially affiliated with the channel)
=MY SOCIAL MEDIA=
Twitter: / rm_transit
Instagram: / rm_transit
Website: rmtransit.com
Substack: reecemartin.substack.com
=ABOUT ME=
Ever wondered why your city's transit just doesn't seem quite up to snuff? RMTransit is here to answer that, and help you open your eyes to all of the different public transportation systems around the world!
Reece (the RM in RMTransit) is an urbanist and public transport critic residing in Toronto, Canada, with the goal of helping the world become more connected through metros, trams, buses, high-speed trains, and all other transport modes.
It was quite noticeable when this switch was made in the Netherlands, and it's still ongoing. E.g. the train between Utrecht and Arnhem used to cross number of rails on both the Utrecht and Arnhem side, and in both cases it now crosses a tunnel, where the other train goes underneath (respectively to Eindhoven from Utrecht, and Nijmegen from Arnhem).
It was a huge undertaking, but it's much better. The trains can approach stations at a much higher speed, they no longer have to waddle their way across the last miles towards the stations.
agree, many European terminal stations with massive track layout have severe speed restrictions approaching the platform, while like I have seen on cab rides, japanese trains can approach the platforms at a much higher speed, and simplified layout also allowed better segregation of platforms for suburban commuter, traditional line intercity and high speed trains
It's seriously agonizing when the last 5 minutes of a train trip is just the train crawling through a mess of switches.
true, but it might also be because they often are much older or there is just not enough space. There are switches you can cross with high speed but the also need more space and i often get the feeling only pretty new built switches are high speed switches@@lsp6032
@@MichaelfromtheGraves Why are the approaches so slow? Do the switches themselves only support slow speeds, or are trains kept slow to reduce the chance of mistake and collision?
@@mindstalk Yes most switches only support crossing them at low speed. High speed switches do exist but they are massive for instance 140 kmph switches are twice the length of 80kmph switches.
You forgot to mention that the Japanese approach is only possible because Japanese rail operators put a tremendous focus on both staff training and preventative maintenance - especially the last item. Stressing out on maintenance means that the risk of both train and infra failures are substantially reduced, which allows operators to cut on redundancy. Because you are sure that there won't be any train or points/signal failures, you can eliminate excess trackage and points which otherwise would be used to "dodge" issues. Staff training is also essential; personnel must be ready and trained to respond and clear issues ASAP when they happen, in this aspect the Japanese also excel, aided by the national trait of as much as possible trying to avoid leaving things to chance.
The film footage showed at least a couple of scenes taken at Hankyu's Umeda terminal in Osaka. This is a classic case of simplifying to the bone and separating fluxes to improve reliability. Tracks in the station are divided into three groups, one for each of the railway's main destinations (L to R - Kobe, Takarazuka + branches and Kyoto + branches). The only connection between the three groups are through two reversing tail tracks, located between the Kobe/Takarazuka and Takarazuka/Kyoto groups. There is no way possible for a train coming from Kyoto to cross into the Kobe group directly for example. Such a move would require a triple reversal, something that Hankyu only does at night.
I think it kind of goes without saying, its another case of less - but better!
As others have somewhat pointed out already, if you can do this, then great. But it requires that you have more than a single doubletrack corridor for commuter, express, "high speed" and freight trains (cough, cough Czechia). And this is why Japan can do it and not many other countries can. Because freight trains absolutely require flexibility, branches, junctions, overtake switches and so on. Japan doesn't have that many freight trains (they have ports - so they don't need them). But most of Europe and NA cannot separate freight from passenger traffic, we don't have the infrastructure, aside from a very limited number of HSRs (again, Japan with it's noodle like shape has a huge advantage). Generally, complex junctions should be replaced with flyovers, but that is expensive and won't fully solve the freight issue anyway. If you, however, have a passenger train only station, then this is a good idea (there are not many of them where I live).
I can only agree, as this is 100% my first fought on this: it all comes crasching down once you have freight traffic alongside the commuter trafic.
Tokyo also has relatively few long distance passenger trains that still use the conventional cape gauge network (ie not the Shinkansen) which helps. I believe the most frequent are the trains to Yamanashi prefecture (which doesn't have a Shinkansen) that mostly use a dedicated island platform at Shinjuku station. I think this is what enabled a lot of simplification work, if you look at old pictures of some of the large stations in Tokyo you can see the track layouts often used to be more complex and sometimes larger, which made sense when these stations not only had to service freight trains but also locomotive hauled long distance and sleeper trains (which have long dwell times and need considerable yard space to build the trains and service them).
Ueno station also has a large-ish number of dead-end platforms for long distance trains on the Tohoku line that I don't think are used much anymore, and these have (or had) a fairly complex station throat that allows trains to access both pairs of Tohoku line tracks and the Joban line from or to any platform, though it may have been simplified as part of the Ueno-Tokyo line project.
In Paris, the RER was created by taking the busiest line of terminal station and liking then by a tunnel. Lot of initial investment but the success cannot be denied. The A line is the world most used outside japan and the B/D common tunnel is the most widely used railway segment.
So if the instra doesn't exist, it's can created from what exist.
I think it is more a question of willingness and finances vs profit. Building rail infrastructure isn't cheap, especially where land acquisition is required. Do not think that Japan does not have the problems that Europe and NA have when building new rail infrastructure. The Odakyu Electric Railway took 20 years overcoming NIMBYs and land acquisitions to expand the busiest section of their mainline from double to quadruple track to support more trains for peak hours. Japan has the willingness to commit the finances and the will to overcome the challenges of building rail infrastructure as they see value in it.
Sounds like it could be a very good option for the northeast corridor, especially around nyc
It’s really funny how this video was posted now, right after it was made public that the agency building railways in Sweden actually doesn’t know the state of many of the switches at big stations, like Gothenburg and Stockholm central 😂
That is . . . concerning
On a separate point, I am glad you looked admiringly at the Netherlands. Its public transport is excellent and in Europe is probably the second best after the Swiss. As most Transport Enthusiasts know (but British politicians do not) , the Netherlands is 'Not Just Bikes'. The Netherlands succeeds as a country because it has an indivisible Trinity of Bikes, Good Public Transport and COMPACT well-planned towns.
NJB sucks, he's too angry and hateful. RMTransit, City Beautiful, and Oh the Urbanity! are the best transit RUclipsrs. I haven't watched much of Alan Fisher and CityNerd, but they're good too.
I like the Netherlands but you think so highly of yourselves. I don't like lists as they are always subjective and easy to manipulate but I'm still interested to know by what criteria you've decided that "in Europe is probably the second best after the Swiss"?
(as for the other two parts of the debatable trinity only the bike infrastructure is something you can certainly declare as one of the best in the world).
@@crowmob-yo6ry I would not speak quite so strongly against NJB. What upsets me about NJB is that he does not seem to know how good Dutch public transport is. NATIONWIDE. Dutch public transport is 'Not Just Trams', (or trains).. Dutch bus services are very good, and are rapidly being electrified.
@@Fan652whe knows it, he’s mentioned it quite a few times on his channel
I'd say the dutch public transport is OK,but nothing to write home about. It got me to where I needed to be fairly quickly. A big minuspoint is the cleanliness, somehow the floor and even the windows are dirty and there's no bistro or restaurant on intercities (at least not on the ones I took, Amsterdam to Groningen). What I found positive were the busses and their credit card payment system, which is really tourist friendly. So I'd say it's in 3rd place in Europe after Switzerland and Austria.
concept of "organization before electronics before concrete" is so simple and beautiful, but in rest of the world it's usually "concrete, concrete, organization and let's put this electronics somewhere it doesn't interupt but doesn't help as well".
Oh for sure, it's still forgotten far too often
what I see at many stations: although having many switches and possible routes across the field of numerous tracks, they try to minimize the number and use of different routes and use the very same routes over and over instead; They still have the option to use another one if really needed - like a backup plan.
Exactly.
And when the alternate routes are used, everything collapses. People have to run from one platform to another because their train was rerouted, the crossover add further delay to trains already delayed.
@@57thorns Which is still orders of magnitude better than the train being stuck out on the line before the station.
@@57thorns not necessarily. Let's take an actual example: Regensburg Central Station. West of the station there's not only one track going diagonally across all platform tracks, which serves both arriving and departing trains from platforms 1, 4, 5 and 7, but there's a set of 2 parallel diagonal going tracks. Thus, if switches on one of the diagonal going tracks fail, trains can still be lead to the intended platforms via the other diagonal going track.
That's what I'm talking about. Having a backup plan to keep things running, if possible in the very same manner as it would be without the switch failure.
Just yesterday a Munich based youtuber published a video from a small station, not sure if it's in the city or not, but that station has a real bottleneck: a freight train had to wait for the commuter train to pass, okay, that's fine. An oncoming freight train had to get onto the other track, but had to wait for the commuter train as well, was then able to proceed, and only upon that the first freight train was able to proceed as well. All 3 trains had to pass the very same section of track as there were no parallel switches, but only switches behind each other. That took them like 10 minutes and several level crossings were closed equally long.
London Bridge redesign is another example of this type of service separation. Bermondsey dive under, plus an additional viaduct to the West of the station means there are now separate lines/platforms for Southern, South-East (Charing X), Thameslink and South-East (Cannon St.), and the removal of a lot of flat crossing points and conflicts
Yep. Unfortunately, on Thameslink at least, as they didn't follow through with terminating Sutton trains at Blackfriars there are still conflicting moves just south of Blackfriars.
It's pretty obvious that it's better at London Bridge but there is trade off. Flexibility does allow for better recovery from problems, The simplest system is a single track but all you need is train to fail to cause serious delays (I've experienced this on the Southminster branch line,
@@eggchipsnbeansthe problem is that the problems then spread to the other branches. There's a happy medium.
I thought of this as well, one breakdown and it all comes to a halt. But then, with the cost savings of building a simpler system, the railway could invest in more reliable Trains@@eggchipsnbeans
@@tonypiccini8475 Well maybe but no train is 100%. In any case, it remains a question of balance; what you gain on the swings, you lose on the roundabouts. I think my point is that I reckon the video is too favourable to the separate lines system.
The wider lesson here is that there's an awful lot of value in "Understand what makes the most successful system good, and copy it." The UK has a persistent problem of copying things from North America that don't work, rather than things from Europe or Asia that do. And that's not just restricted to transport.
that understanding part is very important. Trying to copy without understanding leads to 'cargo cult' like nonsense that just doesn't work, costs the earth, and gives the perfectly sensible way of doing things a bad name in one way or another by association.
The famous Anglo curse
the problem with copying NA is that outside a say NYC, There is no good train cities. And NYC is not the best example because its infrastructure is old and the way public agencies have to contract by law gets them substandard work at way over budget and usually late. I mean it works and I have never had issues taking trains into and around the city but still it needs a lot of upgrades. The bones are good but they need to be upgraded.
He said not just restricted to transport (e.g. car centric city design), for non transport I'd give the creeping privatization of the NHS one example, hopefully that can be stopped in time! @@neoblair-ford3211
@@neoblair-ford3211 I'm also curios.... 🙂
Damn, you really hit the nail on the head here. My journey through engineering was like this:
- 2 years: The system should be able to handle everything
- 4 years: The system should be well documented and tested so we know exactly what the capabilities of the system are and how to operate/maintain it.
- 7 years: The models that go into an engineering project are naturally imperfect. Defining boundaries between what the system can handle and what humans should is critical for avoiding overcomplicating the system as a whole.
- What I expect I'll say at 10 years: Don't overcomplicate your system. Build it to spec, add your margins, and foresee points of failure.
On the other hand, you could argue that the Japanese system needs dedicated lines for each service because the intensity of train service is much higher than on many other systems. Any Tokyo commuter service tends to have trains every few minutes - even the Tokkaido Shinkansen has a train every 4 minutes at peak hours! If you have that many trains, you just have to have dedicated platforms and tracks for them - you just can't intermingle different services. Also, the Japanese emphasis on reliability and extremely high levels of punctuality means that the simple system usually works well and you don't need to make alterations - this is unlike many other countries (like the UK and seemingly now Germany) where punctuality is appalling so you need the flexibility to make alterations or the chaos would be compounded by station delays. Also, don't forget that there are many different train operators in Japan (especially in Tokyo - JR East, plus the two Metro/Subway operators plus many private companies (Keikyu, Tobu, Keisei, Sebu, Tokyu Corp etc etc) all who own their own lines and there isn't a lot of operating over other people's tracks (except for some operators that run through onto the metro lines)..
But high intensity/frequency is a great thing!
@@mindstalk Well of course! But you have to have sufficient infrastructure to handle it. In the UK, Govts are often reluctant to invest in proper railway infrastructure, so what we have is pushed to the limit by the number of trains. But this means that when things go wrong - which they invariably do in the UK and the appalling levels of punctuality, chaos instantly results - there is no slack in the system.
japan also has purpose built lines for trrains that come every 3 hours
Also, a Train leaves Gare de Lyon on the LGV Sud-Est every 3 minutesx despite running on regular non dedicated switching infeastructure, but they still maintain punctuality
Jakarta's subway system is designed after Japanese study, modelled after Singapore. One thing that I note is that they keep saying: the yearly cost of maintaining a narrow gauge track switch is equal to a Lamborghini Huracan a year.
As an amateur train enthusiast, I keep on adding too many switches whenever I play Railroad games like Transport Tycoon, Transport Fever, and the likes. In reality, this indeed causes a lot of complexity and overrunning cost.
While Japanese train system is quite a nightmare and confusing for first time visitor. It kind of started to make sense after I went there for work and do that morning commuting routine through the bustling Tokyo station with confusing amount of interchanges and numerous platforms. All those familiar train services, shinkansen, regional expresses, the Narita express, or the suburban express, or the loop line, will always stop at that particular track - given the amount of people walking by in super busy station, it helps quite a lot, knowing NEX train will always be at the underground platform 1 and 2. Their reliability also gave some degree of assurance and comfort that you will be at your destination at your own time of choice.
That being said:
They have a nasty queue, Japan's mostly manual process means a rather very slow process, despite the tidy line.
Multiple different tracks running through the same area means very little space for crowding passengers, queue often doesn't form up properly during rush hour.
The JR Thunderbird for instance, while having a dedicated platform, they also have a very nasty high speed switches that makes quite uncomfortable ride, especially when nearing Kyoto.
"yearly cost of maintaining a narrow gauge switch is equal to a Lamborghini Huracan a year" is there any paper/source?
Is that means narrow gauge switch are more expensive to maintain than standard gauge? Or vice versa?
@@meongmeong3599 The only reference I could find to those numbers was a paper published in 2001 from the University of South Florida detailing whether narrow-gauge local rail is a viable transportation alternative in Florida. They found a train line (Toonerville Trolley) that cost $150,000 to $200,000 USD per year for a 5.5 mile long train that only runs 2 times a day for 2 months of the year and 2 times a week for another 2 months. Essentially, this line costs that much with less than 200 trips a day. The reason for the high cost was it was fully custom.
That same paper then outlines another track at the Atlanta Zoo which cost less that $15000 USD a year in maintenance, most of which is preventive. Those trains run 8 hours a day everyday of the year except for bad weather on a 3/4 mile track. They even have made profit off of it. It was so much cheaper due to the fact it uses well tested trains and other parts.
So no, a narrow gauge track SWITCH does not equal a Lamborghini Huracan a year... The only reason it should cost that much is if you are a moron who has no idea what they're doing and insists on doing everything custom.
If anyone wants to read the paper its titled "Evaluation of the Economic Viability of Narrow-Gauge Local Rail Systems"
"They have a nasty queue, Japan's mostly manual process means a rather very slow process, despite the tidy line.
Multiple different tracks running through the same area means very little space for crowding passengers, queue often doesn't form up properly during rush hour."
But the thing is that you will not have five different destinations on the same platform, so everyone is going on the next train on one side of the platform or the other.
As long as space is given for those getting off (and you know where the door will be) to get into the center, things will go smoothly.
Compare this to a situation where trains are arriving at different platform depending on delays in the overall system, and people going to the platform just to learn there is another train to depart before theirs.
And you could have two trains leaving in different directions at the same platform practically at the same time (sure, the platforms are "A" and "B", but it is still confusing).
@@AndrewDragon thank you for your information
I found the video very interesting. However, I still would like to point out that Zurich is doing something more traditional, with Zürich HB having all the flexibility in the approach (except for underground platforms). Furthermore, instead of a few simple ü-bahn lines, the city works with a complicated network of many different s-bahn lines. However we all agree that Zürich ÖV is amazing.
My guess would be that the "Japanese way" is much more efficient to achieve throughput (making sense in an very populated area like Japan), but the swiss objective is rather to link as many possible places as possible, with trains being often not even full. So the s-bahn in Zürich guarantees a direct connection from HB to a huge amount of neighboring towns. Achieving the same amount of connectivity in the "Japanese way" would imply a completely over the top network for a tiny city as Zürich.
What are your thoughts on this?
I would think (and just comment in a separate comments) that a complex approach may be a better choice for a stub end terminal. As a stub end approach speed are already rather low, so faster approaches aren’t really a benefit. As trains have to revere out, trains remain at the station for longer too, which usually results in more platforms.
Zurich's system is way more comfortable to ride on, not having to change and not really overcrowded trains is just a better user experience. frequency really isn't the non plus ultra if that means having to change trains and therefore a longer trip. Every heavily used transfer point should be analyzed and interlined.
@@IamTheHolypumpkinThe UK selected simplification for the approch to kings cross, one of the most important teminuses in the country. And the effects have been good. To me that's a vote of confidence in this model. Zurich ÖV is likely effective *despite* of the infrastructure, not because of it.
My thoughts are that the Swiss system while good is likely going to be harder to scale and transfer than the Japanese one. But someone might be able to convince me otherwise.
I would argue that the addition of all the tunnels and flyovers around zurich proves the point of the video.
many of the busiest connections go through the tunnels rather than the overground terminus, where trains very often depart with slight delays.
I lived in Japan for 6 years from 1991-97 and then visited extensively in 2005 (2 week JR pass around the whole country), 2019 and 2023 and have also had layovers at Tokyo Haneda, Nagoya, Osaka and Narita many times. The first time I lived in a suburb of Tokyo and I described the rail system at the time as "an expressway of trains" in my journal. In fact, the line I used to commute to and from work everyday had quadruple tracks and carried, Limited Express, Rapid, Local and subway extension service ALL at the same time. The first couple times I actually missed my station because I caught a Rapid train instead of a local service by mistake. I had never seen 4 different trains running on the same line at the same time - it was INCREDIBLE. No doubt about it - I have been to 106 countries and Japan has the BEST rail system in the world in addition to Tokyo having the best metro system in the world. All others pale by comparison - because not only is it incredibly dense - it is also always on time!! It is very easy to use the system once you get used to its scale. When you arrive at a station on a limited express, it is usually just an opposite platform where you can transfer to the rapid or local service you need. Easy Peasy - Japaneasy!
@BrianBaileyedtech any photos or videos you caught of the trains?
Loads but not from the early 90's as that was before digital cameras and the internet. I will post a few from my trip last month and from 2005's epic two week trip @@longiusaescius2537
It is a truly marvelous transportation system and frankly one of humanity's greatest achievements
Love your channel by the way! I have been a city planning and public transit lover since I was a child.@@RMTransit
Another way of framing this argument is the pros and cons of interlining. Interlining, de-interlining, and track sharing (whether on a main line or at a station) increase complexity, costs, makes automation difficult, and can hurt speed/frequency. But it has its advantages...without extra track capacity, if a train is slow in boarding passengers or turning around (especially if an end-station vs a through-station), the train can end up blocking other trains from entering the station. Trains being able to switch tracks are great for dealing with maintenance and running express routes. Chinese HSR stations are massive with many tracks because of this. But less track sharing means the station can be smaller and less expensive which is good too. Shared lines also in Manhattan or the Chicago loop mean one-seat rides to the downtown office districts are possible without having to do an additional bus/trolley/or metro transfer, but it does cause congestion, speed limiting, and is expensive. It's a tough trade-off.
These concepts and strategies also apply in other industries. For example, software engineering. Famously, in May of 2023 Amazon Prime Video released an engineering blog post about how one of their teams had built a serverless (and highly AWS-componetized) solution based around microarchitecture principals in order to monitor the quality of video streams. This solution gave them massive flexibility to scale up any very specific individual part of the process as needed. However, the AWS components they used only allowed them to reach maybe 5% of the overall performance and resources necessary to monitor their current traffic. They moved toward a more "monolithic" architecture and wound up saving 90% in costs while reducing complexity and improving scale -- ie, performance became their primary concern.
I was also reminded of software engineering, specifically "favour composition over inheritance". Rather than a flexible abstract system with a lot of polymorphic reusability, it is better to have separate independent parts combining to create a greater whole, connected through clearly defined, reliable interfaces.
France is a funny example: SNCF is the opposite of all that, with France's love of infrequent direct trains (and almost all trains have to go to Paris...) while the French way of doing tramways (since the 80s-90s) and metros (since 1900!) is to have single lines with little or no interconnection.
Soviet style of transit was very similar: all metro lines are completely separate from each other (Baku serving as a rare counterexample), whereas the railway for the whole continent is a single mesh of tracks where you can route anything to anywhere at any time. Trams are small and rail-ish.
But still, it seems the idea made its way into the grande maison (sncf). They now emphasise « exploitation en tube » when renovating big stations (ex: lyon part dieu).
I've experienced this in railway simulation games: you always start out with "okay everything can just go anywhere, easy!" but once you're past like 6 platforms and all the trains are waiting for other trains to cross, at some point you go "okay this line is just gonna get its own tracks no one else gets to touch!"
Great analysis of systems philosophies. I spend a lot of time in Tokyo (and teach a seminar on it that includes a week on transit in which I used your vid on explaining Tokyo’s public transit system!) and just got back from a weekend in Toronto, so I could better appreciate your points. Living in hyper-car-centric hellscape that is Nashville, Toronto is public transit heaven to me despite its flaws. But Tokyo is in a league of its own. I simply love the transit there.
Well put. I wanted to bring up something I watched a while ago, a video of a Japanese Shinkansen terminus with 4 platforms and a single point into a double track corridor. Long trains (10cars+) enter and exit the station every 4min respectively (so a train through the point every 2 min), and the controller and drivers operated so smoothly that delays of up to 2min due to slow boarding could be made up without any lasting delays.
An airport model that, admittedly smaller with only 4 vs 10+ platforms, runs very tight (like an airport).
I wouldn't really say its airport like in this case!
I, as a Japanese, thank you for this valuable info. I've ridden U.S. Northeast Corridor and Chicago RTA/Metra, and NOW I understand why we don't know which track my train will arrive (in Boston North station, every single passenger had to ask the employee to verify what we were getting into was the correct train, as the sign did not come up even at the time of boarding), let alone which car stops where along the platform, because the basic philosophy is totally different!! I hear that France is the same way as the U.S. To Japanese, not knowing where the train comes is very unnerving, to be sure, myself included.
While the system does certainly have advantages, i do think that even japan can take it too far sometimes, particularly when it comes to the shinkansen. I was in Japan during the summer and due to rainfall west of osaka and closing of the shinkansen line, the narrow gaugue trains were faster, since the platforms at osaka were unable to turn the shinkansen around in a timely fashion. The approach to kyoto took us 2 hours, at which point we were advised to change to narrow gaugue trains going west. I got to himeji that night at 1am. While countries in the eu are not at risk of simplifying to that extent, its still important to be able to react to situations where an entire line would need to close for whatever reason.
Shinkansen need 15 minutes to turn around. When you are running 16 trains per hour out of 6 platforms (due to lack of space), the margin for error is very slim.
Normally it would not be that bad but that day specifically was the U-turn rush so JR didnt want to cancel any trains.
@@NatPavasant oh yes, 100%. My point here was that by separating lines too much you do lose a lot in extreme events, or accidents, by being unable to use excess capacity on some lines to supplement issues. Japan here doesn't have the possibility to fix this due to gauge differences in track, but you don't want to force yourself into a similar situation if you can avoid it by adding redundancy. (What Reece is arguing for, you need a middle ground)
For all the faults of the German railway network, that is one thing it's actually very good at. Despite all the line closures and switch removals of the 80s to 00s with the goal of privatization, it is still so flexible that almost no matter what line goes down, there is always a way to get trains where they need to go. Of course, this does allow lazy planning for construction, resulting in everything being forced to go over a slow single track line while the mainline is closed for months on end... Oh well.
@@NatPavasant
Even less than that to the north, I watched 16 car trains come in to one of the 2 platforms at Tokyo, unload, get cleaned & stocked etc, and load in under 10mins, and leave on time to the second: with apparently no recovery time [at the terminus].
@@stephenarbon2227 are you sure you time that correctly? It took about 1 minute for the entire 16 trains to come to a stop from starting to enter the platform. Another minute for departure. 6 minutes for cleaning. 2 minutes for alighting and 2 for boarding. That's 12 minutes at the fastest. Tokyo station usually schedule each departure at 20 minutes apart (at least). That's usually 3 minutes buffer between departing and arrival, the aforementioned 12 minutes turnaround time (hence 15 minutes in my original comment) and another 5 just sitting there.
For the JR East Shinkansen at Tokyo station, they run max 16 trains on 4 platforms, so each gets 15 minutes exactly.
A similar lesson was learned in aviation a while back. Four-engine planes used to be standard for trans-oceanic flights. This was because more engines provided more redundancy. As technology improved, the airlines determined that fewer engines were actually more reliable because there were fewer opportunities for failure.
@BB-xx3dv yes, the increased reliability allowed for two larger (and more efficient) engines where four had previously been needed.
Twinjets are not more reliable - they're reliable enough, and much cheaper to maintain. It required significantly increased performance and reliability from a single engine to make ETOPS acceptable as the odds of double failures that wouldn't also become quadruple failures became low enough.
I think the simplicity of the Tokyo area’s track system is actually a natural consequence of it having so many private train networks. These were developed independent of each other and thus where they do meet at the odd station, you might arrive sooner but as you mention, users pay for it with sometimes lengthy transfers on foot to another line. However, some networks such as Tokyu Corporation’s are nicely integrated for the user with simple cross platform transfer between multiple of its line such Toyoko, Shin-Yokohama, Meguro, Oimachi, Denentoshi etc. And at the same time, multiple train companies interline their rolling stock on each other’s lines. Really the most organized and well thought network there is.
Thank you Reece for an excellent video, with which I totally agree. I would particularly emphasize that the more points/switches you have, the more points/switches failures you will have. A points failure just outside a major British station causes chaos lasting several hours!
But on the other hand: if you have a lot of switches, you can route trains around a failed switch.
That's the thing that happens in the Netherlands now: switch failures might be less frequent, but once any of the remaining critical switches fail, you're basically doomed.
Sydney has been on a massive project over the last 20 years or so, 'untangling' train lines... Not to mention the amazing flying junctions near Sydney's central station
It may be a difference between local and long-distance traffic, but DB is just reversing a policy that eliminated frequent switches around lines (allowing trains to switch between the usual two tracks) and will now increase the number of those switches again. Because painful experience shows that not having them means little problems on the line lead to massive network-wide disturbances.
Also, for the same reason, you need to be able to choose different routes for trains to take, so they can go around blockages, be they from accidents or broken tech, maintenance work, or WWII bombs found that need removal (which happens pretty much daily in Germany, though not always close to a rail line), or what have you. Things happen, and in larger networks, things happen a lot. Not being able to cope is a really, really bad thing.
There are currently several major rail corridors with maintenance closures (and probably will be for the next decade, until we've cleaned up all the technical debt accumulated for money-saving reasons), and those typically mean "trains that used to pass this station will now instead stop in this other station, and there's a bus service to bring carry (local) people from one to the other". Just imagine if instead, it was "trains stop and reverse in these two stations, and hundreds of passengers will be transported from one end to the other for every train". That's a nightmare scenario! (Which sometimes does happen, and everybody hates it.)
Also, it occurs to me that not only is size relevant, but Japan is (a) pretty damn linear, and (b) has absolutely no cross-border connections with neighboring countries. Those two probably simplify the problem massively.
My Tokyo commute was multiple trains every day. One day a month was late, usually a few minutes, but longer maybe once a year. The local subways and trains were incredible. The long-haul trains are even better.
Well, the Swiss system is definitely complex and flexible with many services run on the same line. At times they even route InterCity trains over lines that normally only carry local trains. As mentioned in the comments, Germany simplified too much and then paid the price.
Most lines in Japan have about 4-5 tiers of service running on the same line at any given time. Flexibility when needed, but scheduled to minimize last-minute changes.
The K.I.S.S. rule for trains. The more complex a thing is, the more likely it is to fail.
Also I totally agree that countries need to look at what other countries are doing to improve their systems There are so many different aspects that we can improve on not just transit
Hi Reece, Great Video! I'm a permanent Way engineer in Sydney. This is so true! Sydney undertook the "Clearways" project in the early 00's however they only got part way through. The issue was that as the project progressed only in part and you had this Frankenstein's Monster of a railway that was neither flexible or reliable. As large junctions (Ashfield for example) were removed, the adjoining junctions still retained their infrastructure and potential failure points. A failure would occur and then the network would buckle. Project naysayers would then come out of the woodwork and say "When this happened before that trackwork that was pulled out would have prevented the full scale disaster." What I've found is that Day of Operations Teams are looking at their narrow view "How do I solve the right here right now if something goes wrong?" don't look at the big picture. They then rail against the change and when the change occurs and something goes wrong say "Told you so". What needs to happen is higher level leadership and a clear vision that says "We will quarantine the failure to 1 line only. It will be a poor outcome for that line, but the rest of the network will be unaffected." Unfortunately that kind of leadership doesn't exist in most parts of the world. Cheers!
To quantize the changes on station Utrecht: they removed 130 out of 200 switches. It now has 4 times 2 platforms, which are per group as separated as possible: 4 stations under one cover.
And if you need to take a train from a certain platform to the shunting yard - you're out of luck, need to drive 10km to the furthest station to turn the train around, because the switch that would connect the platform track to the shunting yard directly had been removed.
@@marco23pThere is a video here on RUclips by a train driver who had to bring his train from platform 18/19 to platform 5/7 at Utrecht Centraal via Driebergen-Zeist, due to works near Geldermalsen(?). More logical planning would have this train (its whole series) arrive and depart from the same platform. - But then, I don't know the cause for this detour. (Note that the train would also have had to wait somewhere. Time for that parking had been found at Driebergen-Zeist.)
@@marco23p That's not at the expense of the traveler, and may improve stability and predictability. Efficiency is not always the winning argument.
@@markovermeer1394 But it also means that you can't get a train from the yard to a platform track quickly. That would be useful for example, if you want to start a replacement train, but is practically impossible on a short notice nowadays.
Japan is also going along the way of turning junctions into flyovers (Eg Hankyu Awaji, Yamato-Saidaiji), not only for the reasons in the vid, but also to remove level crossings
The Moscow metro is a good example of separation. Every line runs on it's own tracks and stations and only 2 lines have branches. This allows for trains to run even more than once per minute. And it's also easier on passangers, since you can mix up trains arriving at the same place (and makes the map more beautiful)
And impressively they can run under 2 minutes headway in peak hours WITHOUT CBTC! They just use conventional block signalling and fully manual control of the train, comparable to GoA 1
a vast majority if not most of China's metro lines are operated as services on separate tracks as well
@@meongmeong3599 This is not entirely correct. Even though the foundation is conventional track circuits with automatic block signals, all lines and trains employ at least continuous monitoring of next signal state and speed to perform emergency braking, with some lines (actually most of them) supporing more advanced granular speed control based on the block state. In latter case track signals are usually shut off entirely.
No wonder why JNR and eventually JR East had to completely redo all the platforms at Shinjuku Station. They had to to accommodate multiple trains coming in from several train lines at the station itself. In fact, JR East used the tunnel from the abandoned Narita Shinkansen and built elevated tracks to accommodate track expansion at Tokyo Station itself.
In northern India, they have built a dedicated freight corridor capable of carrying electrical stacked trains (overhead wires). You need to grade-separate your traffic into 4 or 5 streams and also separate your passenger and freight junctions. This was too expensive when we started out and traffic was low. Now, it's one hot mess.
It's important to strike a balance. Even in Tokyo, where the busiest JR lines (Yamanote, Keihin-Tohoku, Chuo-Sobu, Saikyo and Keiyo) are more or less segregated, there are still several lines with heavy interlining. They don't weave in and out of each other within central Tokyo, but once you get out past Yokohama, you'll see that the Tokaido Main Line, Yokosuka Line and Shonan-Shinjuku Line all interline with each other. Same to the east past Chiba, where the Keiyo Line and Sobu Rapid Line interline and fan out. There's actually also one split within Tokyo itself: the Joban Line splits from the Ueno-Tokyo Line (Tohoku Main Line) at Ueno, and diverges from the corridor after Nippori.
The Chuo and Sobu lines are also different lines outside Tokyo and the Chuo line has multiple branches and interlines with the a Tokyo metro line.
Sydney is finally starting to learn this lesson - for example, taking the Bankstown line off the city circle and converting it to Metro. It’s so common for one problem to shut down the whole network because all the lines converge
Agreed. We need to keep converting lines to metro to disentangle the system. Ideally we'll end up with one line per track like Tokyo. Eventually the City Circle will be converted too, as it's terribly designed for DDs and we can get rid of the eyesore at Circular Quay
From my experience, the swiss railway network doesn't take the japanese approach at all. We have switches everywhere and thus have a lot of flexibility. So basically the swiss railway are pretty much doing what you said is bad yet you provided Switzerland as an example for how simplification can work for railways...
I visit Switzerland often and know that you have 'switches everywhere'. Correct me if I am wrong, but I think the new layout North-East of Bern Hbf allows all four tracks to be used in both directions. (It certainly allows trains towards Thun and Olten to depart simultaneously.) The advantage is flexibility in operation. The drawbacks are the expense of maintaining all those switches and despite the Taktfahrplan trains on the same line sometimes use different platforms at different times of day. This problem for passengers is mitigated by excellent information screens, and conductors announcing platforms from which connections depart. Use of line numbers also helps visitors. (Dutch Railways do not use them!)
Every video, where a Hankyu train is shown, is a great video.
Very nice and interesting analysis. Thanks.
Thanks for the really informative and helpful video!
Thanks for just summing up my work.
That’s incredible!
I disagree. For places like Japan where you have extremely high density, you can operate your rail network like a metro like you’re describing. For the majority of cities however, while it is useful to have grade separation in places like big stations, it is hardly ever needed to run your infrastructure entirely the simple way of dedicated tracks. You should build according to the demand and frequency of service you’re going to run.
Your main argument seems to not be about infrastructure, but about interlining. I would say interlining is good for intercity services. For example people going to vacation, since they wont care about frequency all that much. But it is bad for local, regional etc rail service, where frequency is crucial.
although Japan has simplicity you mentioned in this video, they have some very complex operating system too, like Sotets-JR-Tokyu-Tokyo Metro through running and with various service pattern there. Kanto railway operator's seems love this complex through running
Meanwhile Kansai through running are much simpler than their east counterparts, many subway lines also have simple operating pattern; stops at all stations
Exactly what I thought. My impression of Japan is that their trains, especially in Tokyo, are incredibly complex, with all the different companies through-running with each other.
Thru running in Japan urban area is needed to prevent overcrowding on the terminal stations. Better keep people on the trains than the platforms.
Although through running means the lines are connected end to end, which is different from having lines side by side with diagonal tracks connecting them
Through running may look complex but essentially it isn't.
After watching many Japanese rail fan video, talk about Great Tokyo area train system. I think through running is very popular in Tokyo is due to
1. History reason: private company not allowed enter inner yamanote line. (with a few exception and JR still JNR when policy in effective) - through running with metro or interchange with yamanote line.
2. Avoid too many passengers interchange at terminus.
3. Provide more turnaround platform: more platforms for turnaround always better.
4. Avoid unused capacity when approaching urban terminal. With through-running, new passengers near urban side terminal can fill the train, those leave the train near urban terminal , and continue their journey in underground.
Great video. Thanks for posting this. Have a nice day.
I think the only system I know of that has got flexibility right is the melbourne tram system. It's not as flexible as the toronto streetcar system (it doesn't have nearly as many junctions at intersections) but it does have a whole stack of routes through the city streets and parallel tracks on adjacent streets. All this leads to a network which can actually run and divert trams in times of disruptions, which are actually really quite frequent. If the network tried to separate the routes from one another, it would mean that when a line like the swanston street goes down due to protesters, it would hold up that entire corridor causing major problems. Instead they can divert trams down other streets, or turn them around because the trams are double ended and there are frequent reversing crossovers on the network. Most of the time networks try to build in flexibility only for it to be badly managed when it's needed, or services are still delayed. Melbourne does a really good job of providing that flexibility to avoid disruptions, without it causing disruptions.
Sydney's rail network has trouble with interlining, but if they built enough tracks to stop the need for interlining on a regular basis, I actually think the layout of the network as it is right now would be pretty good at providing flexibility to stop delays. This is a similar problem to what New york has, it actually has pretty well separated corridors, but because they have too much interlining and there frequencies are too high, you end up with significant problems. High frequencies are not a bad thing, but you can't have flexibility and max frequencies at the same time.
It is interesting to see that Switzerland definitely doesn't follow this approach.
The entire network is still planned with flexibility in mind : most double tracked lines have both tracks usable...in both directions!
So while works might happen on one track, the other can still be operational.
Yeah, to me it seems like switzerland cares more about a robust service that can almost always operate to its fullest, instead of a high performance services, that has trains running every 3 minutes or similar.
I agree with everything here. Just a couple of observations:
Having dedicated lines for services means that they don't all have to have the same gauge - I believe the Shinkansen lines are on a broader gauge than the regular Japanese long distance trains.
Trams that travel on regular roads need more flexibility because there are more things that can disrupt services - car accidents or breakdown., roadworks, etc.
The other thing that can be learned from Japan is the signage. Every line in a city has its own colour, stations are numbered, every (?) train carriage has an electronic display telling you what the next station will be, where the exits are on the platform, which side the doors will open, the time to the rest of the stations on the line. Signage on platforms tells you which carriage you should be in for your preferred exit on your destination station.
Finally, (and a bit off topic here) the transport cards in Japan (IC cards - Suica, Toica, etc.) work in all big cities - you essentially only need one card for all metropolitan train, bus, metro, monorail and tram lines across ALL of Japan.
Germany right now experiences what happens if you reduce complexity at train stations. At the 29th of November 2023, a train took down some overhead wire at the eastern entrance to Hanover Main Station. As Hanover is a central hub of Northern Germany, and with no possibility to route around the defective track, this caused a pile-up of trains throughout all of Northern Germany, all standing in line to go over exactly that piece of track, causing delays of 10 hours and more for hundreds of trains and ten thousands of people.
This is done in Copenhagen now, resulting in longer travel times, more train switches and fewer trains running ... an improvement indeed. 😵💫
It's really noticeable when arriving at Toronto Union, and the train slows to a crawl to navigate the multitude of switches and junctions. It can make the last km of travel feel longer than some of the trips between stations. And makes the start of the journey out from Union also feel less like "and now we're on our way" until minutes later once all the switches have been cleared.
Interesting topic!
You are spot-on. Complexity is a killer. The old KISS principle is ignored so often with expensive results, not only with the infrastructure, but also the trains-
Scotrail's class 385 Hitachi EMUs were fitted with fancy curved windshields that distorted the driver's view, so all had to be replaced with flatter screens.
Crossrail's class 345 were delayed by repeated software problems- remember that trains once ran without on-board computers?
British Rail's Advanced Passenger Train was dogged by complexity and was eventually cancelled, yet the Intercity 125 was based on existing technology and proved to be a game changer.
Well, in defense of the APT, the project was severely underfunded and generally trampled on, with initial negative press coverage contributing to the project's downfall, and despite that the train's issues were completely fixed soon after it was *de*funded, and it proved to be incredibly reliable and comfortable, with the trains and their technology far outlasting the 125. The 125 was created to quickly apply a bandaid, it was 'big and dumb', which did work well for what they needed in for, but the APT was not a failure technologically but rather societally. It was not dogged down by complexity as much as it was simply ahead of its time, and the lack of trust in the project was what killed it.
_remember that trains once ran without on-board computers?_ Yes. They crashed a lot more.
@@MilliBlomPerhaps it could have succeeded with better project direction, but it had too many major changes in design (gas turbine, then electric propulsion) and too much reliance on untried technology (hydrodynamic brakes). A pity, as tilt systems had already been used in Italy and later on the Pendolino.
@@MervynPartinWell, interestingly enough, the pendolino was designed using the APT's patents (sold to Spain). But, I do agree that the amount of changes in the train's design severely inhibited its production.
@@KaiHenningsenMany crashes are due to extraneous factors- signal faults, track faults, etc., but I would also point out that the DLR had a computer controlled crash!
The one thing that I hate about Toronto Union is the lack of dedicated tracks per line honestly. I hope that turning the kitchener/stoufville into a through line will help by reducing the overall line combinations.
It's worth noting, by the way, that when you indicate that planes can go to any gate, this is not actually true. Different gates will be designed to serve different sizes of airplane, and even then, airlines will have specific gates they are allowed to use, and only in very rare circumstances will they be allowed to deviate from that. You generally don't notice this because airlines schedule around it, but in the case of a severe schedule disruption, or if you're flying into an airport like Burbank where the margins are very tight due to the low number of gates, you might have to wait out on the ramp to taxi in, even though there are plenty of other gates open. Aviation does the Japanese model too, it's just done behind the scenes.
I'll be going next month! I'm so hyped
It should be noted that in large cities in Japan, mutual direct rail service is very common. For example, the Minato Mirai Line running through Yokohama can cross up to four companies (Tokyu, Tokyo Metro, Tobu / Seibu) in a row without having to change trains. This means that if a delay occurs somewhere, it can spread to several railroad companies at once. Each company is not lined up like a chain, but rather a complex web of connections that can sometimes cause startling delays in far-flung locations.
Which probably also contributes to the mindset of simple and reliable operations over everything else, you don't want to be the reason for another company's delay
Really good thoughts very insightful, very much in a similar vein to how mixed traffic/speed railways always compromise efficiency for that versatility. The remodelling of Derby Station on the Midland Mainline in the UK is an example of a baby step into this sort of railway design in the UK.
Exactly, and well put. One of the biggest problems with interlining is that a problem on one branch then migrates to the whole network. It also makes scheduling difficult and less robust. London Bridge mainline station was recently rebuilt with minimal conflict between each of the 3 through and 1 terminating groups of services. As you say, get the infrastructure right, and everything else should follow
Grade separation is the way to increase capacity and resilience with of any form of transport (even roads and canals). The less conflicting movements, the better. Having things completely separate in normal operation is obviously the ideal way to achieve perfect grade separation and will alwaus be the aim for civil engineering projects, but...
Unfortunately cost is usually a big constraint and results in less than ideal designs - for example building a 2 track station with 4 lines on the approach or perhaps having a flat junction to avoid building a flyover over another line.
I'd assume that with canals it only really applies if there is a sufficient height difference to begin with, considering that the locks needed to get grade separation otherwise would probabky eat up most of the gains for at least one of the two canals
@@cyri96 It was quite common for busy water ways in the UK to pass over eachother on bridges, even if they weren't at vastly different heights normally. I don't think time was the main concern, rather just not having different routes interfering. In the Victorian era there was *a lot* of work to divert waterways with many drastic course alterations all over - so altering the landscape to alter the elevation wasn't beyond the realm of possibility.
@@izzieb Using horses to pull the boats might have been influencing this as well. Can't really ask the horse to swim or jump the canal you are crossing.
@@57thorns Fair point too.
It is nice when foot/cycle paths and roads are separated. My local government insists that the desire line for a proposed route goes along a busy road and over many intersections instead of the alternative of extending an existing completely separated cycle path to the same destination. (Obviously the real reason is that they're greenwashing resurfacing the roads by larping as pro active transport)
It depends which is less reliable on your particular system....Points/signals or the rolling stock.
In the system you describe,it would only take one train sitting down and the whole job goes up the wall.
In the UK it would be interesting to do the maths on this...
Reminds me of a station I lived near in the UK - Herne Hill. In essence this is a station where one line, a commuter line, crosses another that is part commuter line, part main line to the Kent Coast. There are four platforms. Almost all the time the aforesaid commuter line uses platforms 1 and 4. The other line almost always uses platforms 2 and 3. There is cross-platform interchange between the lines, and trains are, or at least were, timed to make that work. During the rush hour there were some trains that switched from the pure commuter the line and the other line, and vice versa, but I question how important they were.
Yet the track layout lets trains from either line in each direction use either of the two platforms. This is achieved using, if I've counted correctly, 12 sets of points. I remember that once a train using the commuter line derailed there, fortunately at low speed, on points where it blocked all services in both directions on both routes. The redundancy provided by the points led to massive disruption.
I actually didn't know why but i have realised how systems are being rebuilt to be less complex. New lines don't have as much flexibility and connections and before i thought it's a budget problem but now i understand it is actually an improvement.
Re: 6:30 Note that Melbourne's tramway network is also designed to allow a decent amount of service flexibility which is in line with the the type of service our trams provide. In fact, we do not have enough cross-town tram connections.
I don't think this aspect just benefits operating the network but also riding it. A rushed morning commute is helped a great deal if you know (roughly) what platform your train will arrive, even if it is an intercity commute (as are quite common in the netherlands, seen how busy the rotterdam-the hague-amsterdam connection is). Even during the recent reworks of Rotterdam central station i was quite reliably able to roughly guess in what area of the station i would need to be with few surprises.
As someone wo lived in Japan for two years and used the trains regularly, it was certainly a comfort to know that my train would go from the same platform every time, even if I got on the train before or after the one I usually got on. Thinking back, I would have not enjoyed the experience as much if the trains randomly switch tracks, causing us to have to go up and down the stairs to the platforms again.
While mostly operationally true, I think this overlooks one of the points you touched on a previous Tokyo video about through running and the plain importance of operator competence. Most of the criss-crossing issues occur due to historical station format. Europe and North America are full of terminus style stations and infrequent services. The removal of the Tokyu terminal at Shibuya to instead connect with Tokyo Metro is a good recent example of promoting through running.
Recently in in the Greater Tokyo area, they have been linking lines together with projects such as Ueno-Tokyo line, and the East Kanagawa line which has 4 or 5 operators and different signalling (ATS and ATC) and levels of automation depending on portion crossing in and out and consisting of the Sotetsu-JR, Sotetsu-Tokyu and Shin Yokohama line. Even multiple dramatically different Tokyo Metro lines are fed by trains from this one link. Hazawa-Yokohama-Kokudai station in particular has a lot of services from a pair of platforms and there's a pretty crazy service pattern map which sorts them by the colour on the train display. I think it's a really impressive feat that they managed to get that to run on time.
The same is true for the Asakusa and Keikyu line which you also previously noted has a lot of complex running patterns and branching in/out but the Keikyu line can maintain 28 trains per hour in each direction with manual operation of the entire system. They even connect/divide trains in the space of seconds during overtakes.
But per the video, Plan B infrastructure is also quite important. When the Yamanote or Keihin Tohoku go down due to various incidents, they do route trains onto the other's platforms, or even as far as the Yamanote freight line/Saikyo line. But I note one difference in practice between a lot of Europe and Japan is that Japan rarely configures track infrastructure as bidirectional. It almost always stays strictly unidirectional.
Brilliant well-explained video.
The UK does have a number of areas that do operate partially or wholly like this; e.g. approaches to London Bridge, Manchester Piccadilly, Waterloo etc etc, but people don't actually realise this!
This video highlights the importance of knowledge sharing and cooperation as a way to bolster efficiency in infrastructure. The US and Japan have a special relationship, no doubt. But we in the US need to open the door to intensive schooling on how infrastructure is created and cared for. Cue Japan. We need them as teacher now more than ever. It’s in our national best interest to do so.
There is a flip side to the simplicity coin, however. In Copenhagen they’re trying to solve reliability issues through simplification.
First: a new timetable was launched a few days back. This wonderful operational simplification has removed two trains per hour on the western line and introduced slow-stoptrain-only service on the northern line.
Second: a “Japan style” re-arrangement for the Cph Central tracks has been proposed. This wonderful infrastructure simplication will remove all through-routing of trains on the southern line. Instead these will terminate at Cph Central, forcing everybody to change trains, long-walk style, no cross-platform option in sight.
Oh, what bright and simple future!
9:01 "..from the tram networks of Berlin to the stations of Japan..." - 'Hit me with your transport Infrastructure' (Ian Dury and the Blockheads, rare 1979 B-side)
I would like to add that when comparing the flexibility of non-switched tracks/separate lines vs. switched tracks/network, there is one even more flexible transport system - buses. Apart from allowing the operator to route an individual bus on even more micro level utilizing intersections when responding to failures, it also easily allows the operator to have detours to only-this-one-specific-bus. And the ultimate flexibility - cancel bus lines indefinitely on a whim.
Funny that you'd release this video today, when the Hamburg S-Bahn just did something according to this principle a few days ago!
The reworked network just went into effect, and it's majorly different, cutting in half the number of named service patterns to 4 and completely eliminating the reverse branching that used to happen during peak hours. Frequency on the four surviving service patterns is doubled. Not only that, but the operational complexities of multi-system electrical operations are now relegated to only one service instead of several. And all "lost" connections are still reachable through a simple, high-frequency cross-platform interchange at Hamburg Main Station.
It's future-proofed too, with the planned extension with the same odd electrical system as the existing rural service to be added to the other end of the same line as the existing bit, causing no added effects on other lines, and there's capacity left over on the other city-centre core for another line currently under construction.
Oh I loved this one. Could always use some more focus on those extra technical details that make systems work.
I think it‘s also a problem in Germany: In their railways if one train have a probblem, many other trains will delayed, but it‘s i think because the opposite: Too few switches. Maybe it’s because both reasons.
The operation of different Shinkansen services on the same lines is a good example of how to maximise the use of the line. On the Tokaido shinkansen line, three services share it by using the number of stops. The Nozomi is the fastest with the least number of stops. The Hikari is next fastest with an increased number of stops. The Kodama is slowest and stops at every station. The slower services are timed so that they are stopped at stations when the fast trains pass through without reducing speed. With trains at 280 kmh, the timing has to be very precise. To be clear, all the trains on the line have the same maximum speed. The only difference is the number of times that they stop. On the main lines, a similar method is used to allow rapid services to skip most stations. At some stations you have multiple different services which never cross over, and they may be operated by different companies. At Tokyo station, there's multiple subway lines underneath, many main lines and the shinkansen lines. At some stations, these are literally separated by height, with the subway lines underground, main lines at ground level, and the shinkansen elevated above the main lines. By the way, the main lines do carry quite a lot of freight, and some shinkansen services are now carrying limited amounts of high value freight as well (eg live seafood from the fishing port to the Tokyo market).
I like the little location things in the top corner! Also help I can’t sleep because I had a lot of MnMs before bed
London's District Line could really learn from this but probably lacks the space to make any meaningful changes
In September of 2023, an overhead line got damaged at the approach to the Munich main rail station. The entire operation shut down. No local, regional, or long distance trains were able to get to that station. I happened to take the train that day from Frankfurt Airport to Munich Airport. It was quite the adventure trying to figure out how to get to my destination by taking Munich public transport on the outskirts of the city. I can't help but think that with the Japanese model, only one or a few lines would have been affected, instead of shutting down the entire system.
I've always advocated for learning from Japan! This is a great vid! I'd love to see JR E233 series trains on american suburban railways, with a modified version of the double deck green cars for express trains, similar to the new chuo line cars with their wider doors.
Unfortunately most of the Japanese trains are having very weak crashworthiness standard than American, Japanese trains are rarely use anticlimber as safety measures.
Although their preventive safety features and operating procedures are one of the best in thew world i think
With rail corridors it’s a bit likewise to appartments: When they are roomy, every piece of furniture finds its place and it looks good and you have separated spaces or even rooms for every need. But when it’s only a crimped room you need „flexible“ solutions and need to „switch“ spaces from a place for dining to a space for resting at night with retractable bed and really well-designed pieces of furniture with a high grade functionality, fighting with centimeters. For sure in some cases you need custom-built smart solutions that are mor costly than „off-the-shelf-shelves“.
It’s very similar to railway or public transportation infrastructure sometimes, where you mostly do not plan on a white sheet of paper, but it somehow needs to fit in existing urban structures. PT needs to fight for its space, and to increase it where needed, but it’s not always possible or even positive to push an eight-track train line through a city center in order to improve its connectivity.
And the lack of flexibility will take revenge, as it’s mostly not the switches that are the reason for delays and disruptions - there will always be external reasons for these, and these may even increase in the future due to climate -> weather conditions. Yo need the right switches in the proper places to be able to draw good decisions even in cases of special circumstances that may occur daily. Plans are needed for these times and THEN it’s the time to insulate a stretch of track from the rest of the network but also to REROUTE TRAINS as there often are no other options to reroute all the passengers on a highly frequented line to another already crowded train or other means of transportation.
I work at a german PT agency and I see the outcome of a „less switches“ and „reducing flexibility“ politics that DB NETZ followed a long time every day. I know their approach was not one done in a Japanese manner, but rather cost cutting - but I guess mist people see Tokyo and it’s rail systems and think they have seen Japan and ask why this is not copied in GER. These are probably the same people fighting for an ICE stop in their town of 10.000 in GER. Something doesn’t fit there.
My knowledge of what I am about to say is limited to a half remembered documentary, but I think the Moscow metro is run like this, with each line having its own dedicated tracks and platforms. Different lines don’t intersect. I can see how the operational overhead scales linearly with a system like this.
Thanks for this video, this was definitely quite counter-intuitive for me. My guess is that adding flexibility might lead to building fewer tracks than necessary, which you can't really do if you have completely separate networks.
Okay, I learned something new today about transit
5:57 Hankyu Railways 🥰
Was just in Utrecht for work, the central station was great also got to check out the historic canal that used to be a highway from 70s till recently and the railway museum!
Every video about railroads: Foamers that have never worked for a railroad explaining what the railroads need to do.
They also have that style of operation in quite a few Australian cities but still can do things like change the running through platforms of lines in the event of scheduled works or an incident! :)
One nice thing about the Montreal subway is that none of the tracks are connected except by a few non-service spurs. So few ways for the tracks to be problems.
I used The Japan railway system and it's incredible.
On the other hand you have Grand Central which has to be complex, Multiple commuter lines have already merged farther north into four tracks before even entering Manhattan and then has to split into 44 platforms all underground. I mean I cannot imagine how many hundreds of billions it would take to simplify it and bring Hudson, Harlem and New Haven lines into GCT as their own things. Probably not helped by being structurally 110 years old.
Thanks again for the introduction to Japanese research and analysis. You still have enough dining car sleeper cars over there, don't you? And I have a longing for the friendly streetcars and the general elegance of the EU.
I can confidently recommend the Hankyu's simple paint job, elegant door opening and closing, and magnificent Umeda station.
I'd definitely prefer not to have to change trains to get to my destination, though, after seeing Japan's skinny route network. I could use the exercise and hydration. And fewer freight trains, too. Trucks arrive while containers are being repacked.
Tokyo rail systems are unique among Japanese cities. Smaller cities typically only have a single or a couple of lines at each station and service is usually not that frequent.
Tokyo has several rail companies serving the area with separate fare collection, so each companies' rail stations are physically separated from each other and passengers have to go through gates to access other rail lines. Then, there are so many commuters in Tokyo, the platforms can't handle the number of people that would crowd the platform during rush hour if too many lines used the same platforms. As it is, some platforms are a complete zoo during rush hour.
I saw the same thing on the harz narrow gauge in germany. Where branches came to the station, the platforms were located on the branch line track.
The inside double slipswitch is among the most compact and elegant mechanical solutions ever devised. An inside slipswitch (with the outside points located between the outside frogs: where the tracks separate at either end) can accommodate fairly high speed on the diverging tracks with radii maxing-out at about 1750' - without spiraling. With, a 2000' radius is possible, given a crossing angle about 3.1°. It's a very useful solution, and one you won't find on any other form of guideway transit, ever - perfected over a hundred years or so up to mid-century.
Truly, over-complication and duplicativeness is crucially necessary to avoid, but it doesn't mean the great stations of the 20th Century can be bettered as far as their track layout goes. After all, train control is vastly more efficient and likely to be failsafe now than it ever was before. The current trend in the US is to deliberately simplify to the point of destruction. You're likely to see it soon at such golden age stations as Penn in NY and Washington, DC. It is planned vandalism, at a ridiculous price.
I work in software development. There we have a principle called yagni, "you ain't gonna need it". Simply put, keep things as simple as possible until you run into the actual need for the added complexity
Great video once again highlight the convenience of having dedicated platforms for each line at bottle neck junctions. I love being able to walk up to a platform ahead of a train's arrival with out having to consult the schedule board. It would be great to have GO adopt this solution at Union Station and much like JR, paint the trains in a livery that match the line color as is done by the JR Lines in Tokyo. I am curious how you would layout the dedicated platforms at Union Station based on the GO Lines to facilitate cross transfer - eg would you have an incoming Milton line train share the platform with a Stouffville bound train departing in the same direction?
After seeing this video, I started to understand on the pattern implemented to the several railway station enlargement in Jakarta. Turns out they are starting to segregate tracks to dedicated train lines, by adding more tracks and reducimg direct track redundancy.
This only works if you don't have to worry about disruptions that are beyond your control. For instance, in a tram system, having a flexible network is really needed so that you can bypass a traffic accident or a fire that is blocking one line. And at the same time, it enables higher spatial density of service.