The rise of automated metro systems? A French return on experience

Photo Francois Gerin / Deputy Managing Director, Siemens SAS / January 5th, 2012

The first automated metro lines opened over thirty years ago. Today, new projects are being launched and older lines are being upgraded to automatic systems. These choices are driven by technical as well as economic reasons. But do the edges of the automated metro live up to the investments? Recent projects in and around Paris provide some valuable feedback.

ParisTech Review – Where do automated metros come from?

François Gerin - Though some semi-automated metros have been developed in the US as soon as 1959, the first fully automatic project emerged at the end of the 60s, when Matra Transport designed the VAL metro system to connect Villeneuve d’Asq with Lille, in the north of France. VAL was a revolutionary system, much ahead of its time. Other cities took on the same system, which was eventually implemented on an international scale, in cities such as Chicago, Turin or Seoul. At the same time, other companies started to develop automated metros: Bombardier in Canada designed the automated line which connects JFK airport to New York City. However, there are few industrial players in this field, as it requires very specialized capabilities in an uncertain market.

The first automated metros were light, which means they were of relatively small size (breadth and length of trains). In October 1998, as Matra Transport was taken over by German company Siemens, it world-premiered the first “heavy” automated metro line: line 14 of the Paris Metro, still using rubber tyres, had its train size adapted to the needs of a capital city. The main advantage of such a model is the superior load capacity. The service opened on daily basis of 150 000 passengers. Today, it is working with 500 000 daily passengers without showing any signs of collapse.

Automated metros also show clear edges in terms of liability and capacity, for priority and loaded links. That’s why Siemens’ technology has been chosen by big cities such as Barcelona and Sao Paolo. Automated transportation systems are also used in airports, although for short distances, usually no more than a few kilometres. Metropolitan systems cover several dozen kilometres, which implies many technical constraints as well as completely different ways of regulating speed or the number of trains.

There aren’t many automated metros: hardly thirty in the whole world. However, new projects are being designed and will certainly emerge very quickly because they respond with appropriate solutions to the demands of cities of the future.

What exactly do the operators wishing to implement these systems ask for?

One advantage is to be able to run circulate continuously, without requiring much staff – although human surveillance at distance is still necessary, of course. But the most striking advantage is probably the flexibility to adapt use according to demand. It is very easy indeed to increase the frequency (that is, the number of passengers per hour) without need for additional staff. Besides, as the trains are completely closed and isolated from the outside platforms, there is absolutely no danger of objects (or even persons) falling on the railway. Automation also significantly improves driving conditions, speed modulation and automated emergency procedures. Therefore, frequency can be increased without compromising neither safety, nor security.

For lines which are collapsed in normal time, automated metros offer features than can convince operators to transform traditional lines into automated lines. That’s exactly what RATP is doing on line 1 of the Parisian metro, which currently receives up to 700 000 visitors every day. In this case, automation will reduce significantly the delay between two trains, which will naturally increase capacity as well as improving frequency, regularity and liability.

Optimum speed management leads to better performances as well as energy savings (10% to 15% savings on the total electricity bills). Automated driving is more regular and reduces stress on parts and materials. The energy saved on braking a train can help another train to start. Other improvements are to be counted: greater frequency enables to optimize train occupation (trains are gangwayed throughout their whole length), automated platform edge doors limit the risk of accidents on the railway, thus enabling greater speed in connecting stations. In fact, this last innovation has also been copied in traditional crowded stations.

On the whole, trains circulate faster and more smoothly. They can also circulate more frequently and passenger comfort is greatly improved. During rush hours on line 1, trains circulate every 95 seconds, a performance by far superior to that of classical lines.

Let’s not forget that on crowded lines, drivers also undergo heavy stress. This explains why unions, usually suspicious about job conversions linked to technological transformations, have accepted quickly to discuss and find solutions.

Does automating a metro line generate any problems?

Yes, indeed. And although we could foresee some of them, today’s experience gives us a precise insight of these problems. Take line 1, for instance, which is also the oldest line of the Paris network. Although most of its path is in straight lines, there are also some sharp curves. Our engineers had to cope with it: this implies a very precise management of deceleration and acceleration when entering and exiting curves. The issue of passenger stability had been taken in account by the former line’s constructors, but only for manual drive. Today’s engineers have to find solutions for a framework that wasn’t designed at a start for the possibilities of automated metro.

But the transformation of a classical line also raises operational problems. Setting up new trains and equipment might heavily disrupt train circulation. Automating such a loaded line was far from easy and eventually became a challenge. They had to work by night to make the first tests without passengers. Even if the first automated trains circulate since November 3rd, both systems, classical and automated will coexist during a year. That’s a technical challenge of its own right!

Let’s talk about the technical aspects. Can we draw a clear line between the machines’ management and the operating system, between hardware and software?
Both are closely tied, although the core of our system lies within the operating system that manages the line. In any case, that’s what we work on first, for any given project. The machines’ performance is a rather well-known aspect that serves as a guideline of the scope of possibilities. On a certain project, we might only be requested to design the operating system whereas other competitors are requested to deliver materials and machines.

Although system design is our core activity, other aspects, such as emergency braking systems, need a minute work. The same problem arises when dealing with high speed trains, but in Siemens, as in other firms, this issue is dealt with by a completely different department.

Machines evolve: up to which point do these upgrades modify the design of the operating system?

The whole model might be reorganised – it’s not only about changing the materials used to construct the trains. Take the following example. Thirty years ago, when the first VAL trains started running, the tunnel cost was still high and the trains were only 2.10 metres wide. Today, tunnel digging has considerably improved and we were able to make 2.80 metres wide trains. This evolution doesn’t only impact the passenger capacity, it also deeply modifies the engineering behind the scene: with this new width standard, they can use a central rail guiding system. The whole train’s performance and management parameters are completely modified.

This also gives very different possibilities in the way train cars are linked and thus permits new standards in the length of the trains: 2 cars in off-peak, up to 8 cars in rush-hour. That way, smarter energy and load distribution enables greater frequency, at lesser electric cost. The light metro project between Orly and Versailles (Paris area) we are working on today would offer trains every 2 ½ minutes, which is of great comfort for the passengers, in terms of frequency.

Much ink has been spilt already on this project: how is it that, in a semi-urban environment and on a 35 km line, you offer a light metro when heavy solutions were expected?

For both economic and technical reasons. The light automated metro can climb steeper grades, up to 10% and 12%. Heavy metro is limited to 5%. With the light metro, on a rough terrain we can stay closer to the surface. Thus, we can significantly reduce costs (from 3 billion euro to less than 2 billion) and plan more stops.

One of the issues was the duration of works: with a conventional heavy metro, buried deeper into the ground, a tunnel boring machine would advance some 3 km a year and enables only 2 machines to work at the same time. On a 35km distance, we would need at least 5 years to simply dig the tunnel. With a light metro system, we can combine underground with surface and air: several builders can work at the same time on different parts of the path and work quicker.

Instead of a single, deeply buried station, we can offer 3 different stations to which most passengers can access by walking. These different elements, combined with increased frequency of trains, represent a significant difference in terms of use: they integrate the project in the urban network.

Other options are also possible: even a single bus line was evoked. Projects are selected according to investment and management costs. Although the light metro is relatively cheap in terms of management costs, it certainly calls for a heavy investment. But it’s also about investing in the long-term: if the issue is to interconnect schools, universities, labs, industries and residential areas inside a city in growth, then, the automated Metro offers very clear edges. One of the main issues of this project was to give life to the campus. That’s certainly not a straightforward idea in a French context, which lacks that kind of culture. Making the campus live and connecting it to Paris. If we want to convince researchers and foreign students to live on campus, it’s certainly worth offering quality transport: I don’t know whether you’ve ever waited for a bus at 11pm, on a rainy day, but it certainly isn’t what American students dream of, when spending a semester in France.

What’s the economic model of this type of project?

Each project has a model of its own. But on the whole, they rely on a variety of partners which take in charge specific aspects, requiring different skills. Beyond precise configurations (partnership types, the respective part between public and private investors, choice between fixed rents or operating concessions), we can also question the meaning of these investments. This asks for being discussed, but that’s precisely why a public debate was organised around this project. The debate should help to trigger ideas beyond the interests of organised players, who have expressed their opinions so far. The idea is to enlarge our representation.

For instance, we could imagine a hybrid model mixing passenger transport and freight transport. It isn’t a major issue from a technical point of view: it’s only about installing standard cargo containers, as used in the airports, on the metro cars. We could imagine this system working during the day (without disrupting the passenger traffic), or at night, or even with freight cars at the end of the trains, by isolating a part of the station platform. We would need to imagine an efficient system for loading and unloading goods as well as transportation system for the last kilometre (for instance, with electric trucks). Local players are very much interested on developing alternative transportations of goods in urban areas.

Again, it’s all a matter of costs. It’s also an economic model to explore within a very different transport framework, aimed at sustainable growth. This is a milestone project which implies a great deal of innovation and experimentation.

It is often said that transport is driven by the offer, more than the demand. In this peculiar case, we can’t really measure the demand, which changes according to the offer. At Rennes for instance, the automated line was designed for a daily 70 000 passengers traffic. Today, it’s used by some 120 000 passengers. The “modal shift” – that is, the change from one transport mode to another – has gone far beyond previsions. That’s what we can expect from Saclay: an ambitious offer that might revolutionize habits. It’s not an easy task to make economic measures. For instance, easing the contact between researchers and labs is without any doubt, useful. But how can we measure its economic impact? At this point, we need to shift from economic evaluation, where industrial players and managers have their word to say, and public decisions, which are the responsibility of institutions and political players.

Note from the editors: Siemens SAS is a patron of ParisTech Review.


  • The Future of Automated Freight Transport: Concepts, Design And Implementation

    List Price: EUR 98,83
  • A history of automated metros, starting with the semi-automated experiments

More on paristech review

By the author

  • The rise of automated metro systems? A French return on experienceon January 5th, 2012
  • Pingback: Skynet « Three Leaves