I have just been in Paris for a few days. I hadn't been there since 1994, which was my motivation to hop on the Eurostar and have a look round. Of course there are lots of transport related things I can talk about, not least the Eurostar itself, which I may get to in another post. One thing that is interesting though is to look at the differences between the Paris Metro and the London underground. In particular, the lines in Paris are much simpler, with the vast majority of them containing no branches, and simply going from one point to another. (See this map. Note that the thicker lines denoted with letters A, B, C, D are the RER - essentially an express metro service connecting to regional railways - which are I am not discussing right now). The two lines that do contain branches just contain simple forks towards the ends of the lines.
By comparison, the London Underground is much more complex. (The current version of Harry Beck's iconic map is here). A large portion of it - the District, Metropolitan, Circle, and Hammersmith and City lines, as well as part of the Jubilee line - consists of interconnected lines with various branches and routes, and there are issues with interoperability of equipment and signalling. (The Northern line, with its bifurcations and branches, is similar). The District line contains a hodge-podge of different services with different starts and destinations.
There are of course historical and political reasons for this. The London Underground was designed for suburban services, as well as for carrying people around central London. The Paris Metro was orginally designed only for carrying people around the actual political city of Paris (shown in white on the map linked to above), which is quite small. RATP, the operator of the Metro, was for a long time only allowed to build services inside Paris itself, and it built the Metro in such a way that there is nowhere inside Paris itself that isn't close to a Metro station. The strategy of building single lines without branches that traverse gentle curves solved this problem quite effectively. (The restriction was eventually abolished, and Metro lines were extended a little way into the suburbs, but suburban travel is generally done on a separate set of suburban railways, now named the Transilien). This means that transport is great inside Paris but less good once you are technically in the suburbs. (This practice of treating Paris with great respect but giving significantly lesser treatment to the suburbs stretches way beyond transport).
In any event, the simplicity of Parisian Metro Lines means that they are largely self contained. Each operates simply by trains going from one end to the other, and then going in the opposite direction, over and over. If you are building a new line, you generally do not have to deal with legacy issues to keep it compatible with the systems running on existing lines, but can use the latest technology everywhere on the line. This has been done. Line 14 of the Paris Metro, also called meteor, was opened in 1998, and was at the time state of the art in every way. It is completely automated and the trains do not have drivers, instead being controlled by computers in a central control room. This allows for the trains to run at high speeds by metro standards (they run at around 40km/h compared to 20km/h for most lines). It also allows very high frequencies at peak hours, with trains being separated by as little as 70 seconds. As there is no driver and no driver's compartment, passengers are able to sit at the very front of the train. The front of the train consists of a transparent window, so that passengers can look out the front of the train through the (lit) tunnel, seeing the stations ahead and trains coming in the opposite direction. Plus, the same technology that has led to so called "bendy buses" has been used in the construction of the trains. Rather than having separate enclosed carriages connected by doors, the entire train consists of one single enclosure, and passengers can look from one end of the train to the other. At the end of each "carriage", there is a section of rubber floor, roof, and walls that is extremely flexible (obviously with some metal connections as well under the floor). This allows the train to bend as it goes around corners, without requiring it to be divided into traditional carriages. There are some pictures here.
My simple observation is that the new line is extremely convenient, and if you sit up the front and look at the window, riding it is cool.
Compare this briefly with the most recent "new" line of the London underground, the Jubilee line. This consists of a new section of line going from Stratford via Docklands and Waterloo to Green Park, connected to a section of line built under London to Baker Street in the 1970s, connected to a century old branch of the old Metropolitan Railway. Certainly the legacies of the these various sections of older track contribute to the difficulty of running the Jubilee line, which is not nearly as automated as is meteor. There have been signaling related problems achieving frequencies on the Jubilee line anything like what can be achieved by meteor. The nature of the London Underground is that it is much more complicated than the Paris metro, and due to this, it is very difficult to make even new parts of it truly state of the art. Because the Paris metro is simpler, it is easier.
Crozier visions
I suppose that to many Samizdata readers the quotes below will be old news. But it was newsworthy news to Patrick Crozier when he wrote it, and it was news to me when I read it about two days ago. I realise that two days in blog time is a lifetime, but...
Samizdata.net on May 28, 2003
Comments
Permalink
I rode Line 14 a couple of years ago and I have to say I absolutely agree with you: it is quite something.
On the question of simplicity, I am not sure if I have made the point when discussing Japanese railways but certainly when you make the effort (and spend the money) to make things simple it tends to have some quite marvellous side-effects in terms of punctuality and reliability.
Posted by Patrick Crozier on April 10, 2003