Automatic train operation – beyond the city

From metro systems to mainline networks, Passion4Transport talks driverless rail with Generoso Immediato, product planning & associated technologies referent, Ansaldo STS.


P4T: Why is Copenhagen’s automatic metro – lines M1 and M2 delivered by Ansaldo STS in 2002 – widely considered one of the best in the world?
GI: In my opinion, the success of this GoA4 system [Grade of Automation 4, or Unattended Train Operation, UTO] is down to a combination of factors, namely:

-24/7 service was introduced in March 2009, making it the first, if not the only 100% driverless system operating around the clock within a two-track network. This means maintenance can be performed on one track, while service continues uninterrupted on the other

-the two lines, covering a total of 21km and serving 22 stations, take high passenger volumes (reaching ridership of 57 million in 2015) from Copenhagen airport, across the city centre, through areas of new urban development, and on into the densely populated suburbs

-the overall system, with all stations and vehicles unmanned, has been designed to meet high operational performance. Service availability improves year on year, largely due to a minimum two-minute headway at peak times and maximum 15-minute during the night service, together with an average commercial speed of 38km/hr (max. 80km/hr).


Concerning safety, there are a few key points worth mentioning. From its conception until the start of revenue service in 2002, Copenhagen Metro underwent a thorough safety certification process in compliance with the CENELEC suite of standards for rail transport.

To sum up, EN 5012x focuses on the development of reliable and safe (dependable) railway systems and related components. Its methodological approach has general validity and is widely used for metro safety certification. Moreover, NFPA 130 has been used as an important reference standard for fire issues in tunnels, as well as for the following aspects: sizing and control of emergency ventilation; layout of safety-related cables for power supply and control; other systems for evacuation and rescue.

Figures from UITP’s statistics brief (July 2016) show there are 53 fully automated metro lines in 36 cities around the world. They are operating over a total of 789km; a 17% increase in kilometres compared to 2014 figures



P4T: What are the main challenges of taking Automatic Train Operation (ATO) out of the city (metro) and into a wider (mainline) network context? Are they more regulatory than technical?

GI: Considering that ATO functions at a certain level of abstraction, and technical and regulatory issues aside, I think the ‘key equation’ is the overall ‘degree of automation’ required to address a specific transportation system, to meet passengers’ needs and deliver them benefits, i.e. first and foremost service safety and security, followed by punctuality and reliability, and, last but not least, comfort’.

Over the long term, it’s likely that more than one solution will emerge for the different metro, mainline, and high-speed rail applications. Beyond the metro context, we can assume the three main technical constraints of ATO are complexity, the variability of routes, and heterogeneous vehicle fleets (different types of train and coach configurations).

In addition to Copenhagen’s M1/M2 lines, other GoA4 metros by Ansaldo STS active today are Milan L5 (M5) / 12.9km; MetroBus Brescia / 13.7km; Rome Line C / 18.1km (complete line); and Riyadh Princess Noura University Campus / 11.3km

P4T: The key strengths of ATO for metro systems are generally considered to be shorter headways through time-optimised travel; accurate stopping and automatic door opening; the stipulation of a target dwell time for the driver; and precise travel along ETCS brake supervision curves. Would all these benefits also apply to automatic mainline and regional rail services?

GI: It certainly could be possible. I’d also like to mention two further features of interest:

-‘DAS [Driver Advisory System]/ATO’ – trains are driven according to optimum speed profiles that minimise energy consumption

-all the functionalities supporting passenger comfort and service information. For instance, the ‘ATO [Passenger Information System]/PIS provides visual and audio information for passengers in normal operation and in the case of disruptions. It’s true that in the context of metro service, especially for GoA4 systems, this set of features is perceived positively by passengers, thus increasing their overall journey satisfaction.

On the one hand, increasing capacity and punctuality can be considered direct goals of ATO; on the other, the additional features mentioned above offer more ‘indirect’ benefits.


P4T: The use of ATO over ETCS – could you explain the debate?

GI: This issue is nothing new. In fact it has often been suggested that ATO could be provided using ERTMS/ETCS as the underlying Automatic Train Protection (ATP) system.

Since ETCS was designed primarily in the 1990s for train protection (Class 1 functions) on national mainlines and high-speed lines, the needs of ATO were not taken into account when specifying the ETCS functionality. Over the intervening years, passenger transport needs have evolved, i.e. higher capacity demand, more stringent safety requirements, improved performance of systems, better service efficiency, punctuality and passenger comfort, while decreasing operating costs and energy consumption have become musts. All these new goals make the introduction of ATO one of the major game changers in the evolving ERTMS/ETCS roadmap.

As mentioned during the Rail Forum Europe conference on ‘Automation of rail vehicles and safety: beyond the concept?’ (May 2016), ATO was a relevant research work stream within the TEN-T project (2011-2015). It led to the creation of an ATO UNISIG working group to develop specifications for ATO over ETCS.

From the conference this May I recall both the positive feedback on the results achieved so far, as well mention of the open points regarding the work still to be done.  Don’t forget that the results of the TEN-T project (GoA2 concepts and requirements) will be carried out by the Shift2Rail programme in order to provide a ready-to-deploy GoA2 solution for railway applications.

Moving forward, Shift2Rail represents ‘fertile ground’ for ATO – the latter forms the framework of its Innovation Programme 2 (IP2) The context set for the ATO technical demonstrator will be an opportunity to reuse the key results achieved so far, to gain experience in all railway market segments – mainline/high-speed, urban/suburban, regional and freight lines – and so challenge the ATO concept, as well as developing and validating a ‘standard’ ATO up to GoA3/GoA4 and over ETCS.

In 2011, Metroselskabet awarded Ansaldo STS the contract to deliver Copenhagen’s CityRingen metro – a fully automated, driverless system based on CBTC technology. The project, currently in the development phase, involves building and equipping lines M3 and M4 to encircle the city. Features of note include:

-two single track tunnels each approximately 19km in length

-1 over- and 18 underground stations with island platforms

-4 crossover facilities

-5 construction and ventilation shafts

-an automated control and maintenance centre (CMC)


When in commercial service, CityRingen’s 39 (3-car) passenger trains will run  independently of the existing M1/M2 Metro. Moreover, the system will operate 24/7, with a planned headway of approx. 100 seconds at peak times in the central area (where the M3 overlaps the M4)

P4T: Driverless rail operations are considered safer than manual. Is the technology really more reliable than the human factor?

GI: In our sector of activity, the human factor is always just around the corner. In this particular scenario, the automation of certain functions (de-coupled by definition from the vital ATP application), backed by mature procedures and supervision by expert and qualified operators, could definitely improve the overall safety and security levels of the transport system. It could well reduce/prevent risks, manage, both efficiently and effectively, any situations threatening passengers and staff,  and support/restore the operations in the case of any malfunctioning.


Speaking to German newspaper Frankfurter Allgemeine Zeitung in June 2016, Rüdiger Grube, CEO,  Deutsche Bahn (DB), said the company is planning to introduce driverless trains by as early as 2021

P4T: Is mainline rail automation the only realistic solution for coping with rising passenger volumes in the years to come, as well as expectations for greater system reliability and safety?

GI: It certainly represents a viable solution for boosting capacity and punctuality, as well as responding to passengers’ expectations for greater comfort. Furthermore there is the potential for delivering overall service flexibility and ensuring traffic management and system recovery capability in the case of disruptions.

Combining the strengths of these ATO principles creates a compelling value proposition for customers and passengers alike – first in terms of safety, followed by energy efficiency, i.e. power supply savings, and optimisation of O&M [Operation & Maintenance] costs.

Before reaching ‘extreme’ passenger volume scenarios, which certainly cannot be excluded given rising numbers, the rail industry and operators should together be able to improve the rail offer, which society and passengers expect to be more competitive in the future.

Overview of UTO CBTC turnkey projects by Ansaldo STS

Taipei Circular Line / 15.4km

Riyadh Metro Line 3 / 40.7km

Milan Line 4 (M4) / 15.2km

Lima Line 2-4 / 35km

Thessaloniki / 9.5km


P4T: At the Rail Forum Europe conference, Pierre Izard from SNCF (French Railways) said: “ATO is not a dream but a major project. Train automation should be integrated with the European railway roadmap. Here success calls for collaboration between all the European stakeholders and the strong involvement of ERA”. Why is working together necessary?

GI: ATO over ETCS certainly can’t be pursued without strong teamwork between the parties concerned. In this spirit comes the suggestion by Mr Izard, underlying the need to set a common vision and roadmap for conventional line automation. It recalls the position of ERA [European Railway Agency], which is establishing a framework for defining the next ATO roadmap that includes the following guidelines:


-ensuring interoperability and that ATO functions are functionally decoupled from ETCS

-ATO usable for urban rail, high speed, rail freight and mixed traffic lines

-align EU-funded research activities

-use existing experience/information as input (note: during the May conference, Mr Izard presented real cases of ATO applications by SNCF

What we are debating here is how the rail sector is going to extend the ‘automation concept’. Here, rather than the technology itself, it’s more a question of projecting the rail automation strategy towards a broader and more complex frontier (mainline operation). In my opinion, achieving this goal will call for a wide system perspective and approach, together with strategic thinking.


Photo: Deutsche Bahn AG

P4T: In the road transport sector (trucks, cars), could workforce savings and greater efficiency gained through automation pose a threat to the competitiveness of rail in the years to come?

GI: If we look at the bigger transport picture, I believe rail will continue to serve mass transport demand. Consequently I don’t see any threat coming from road transport. On the contrary, the mode may well prove complementary in certain cases, e.g. rail freight lines and trucks in perfect combination for modal interchange. Here truck automation could really improve its efficiency and cost optimisation; rail freight routes could likewise benefit by applying the ATO concept.


 Photo: Daimler

P4T: This June 2016, Dutch rail infrastructure manager ProRail announced plans to trial driverless freight trains on the Betuweroute route between Rotterdam and Germany. Any comment?

GI: This news suggests that the freight segment does indeed represent a feasible path for further developing rail automation Such pilots are generally a good sign, indicating investment and technical progress. Note, in the Czech Republic, operator AZD has been operating some main lines  with ATO for the past 25 years.

P4T: Automation in transport was a major topic at the TRA 2016 Conference in Warsaw. What are your takeaways?

GI: I attended interesting debates on automation trends, which mostly concern the emerging growth of autonomous cars and trucks. Technologies like Big Data and the IoT [Internet of Things], plus connectivity, i.e. telecom infrastructure and services, all play an important role in this vision, as do clear and sustainable R&D&I investment plans for the near future.

Also with me in the plenary sessions were other colleagues/experts from the rail domain. We brought our sector experience to the table, explaining how automation in rail – especially the metro application – is a reality, and underlying the fact that it is now a  mature technology. Today, dozens of fully automated metro systems are  efficiently serving millions of passengers every day, in some of the biggest and most congested cities across the globe.

My personal conclusion to these debates in Warsaw are that technology is a means; automation must be seen as on opportunity to satisfy needs, gain benefits, and minimise threats and risks; a ‘system vision’ is key to bringing everything together.

An expert in Unattended Driverless Metro Systems, Generoso Immediato has 10 years’ experience in the railway sector, working in the fields of project execution and product strategy, as well as R&D&I management. LinkedIn / Twitter – @genimmediato





Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s