From Glasgow to Dubai, Algiers to São Paolo, Automatic Train Operation (ATO) is in commercial service on 52 metro lines in 32 cities across the globe. Boosting capacity, punctuality and safety, eliminating routine operations, providing computerised support for decision making, and reducing maintenance costs, the technology promises a host of significant benefits.

Little wonder main line rail is looking at driverless operation with a favourable eye. Essentially, operators want the automatic functions for the following reasons: to help them deliver improved performance for varied service patterns on mixed- traffic networks, to optimise speed curves to save energy and to reduce wear on both trains and the tracks.

Yet taking ATO out of the city and into a wider network context is no easy task. To address the issues, Rail Forum Europe (RFE) recently hosted a meeting in Brussels to bring the related technology, regulatory, and social implications to the table.

‘Local’ or ‘network’ approach

Driverless main line rail is problematic due to its diversity of trains, network complexity, and the presence of multiple operators. “The railways are fragmented and they function as networks,” said Josef Doppelbauer, executive director of the European Railway Agency (ERA). “So when innovating, when introducing a new technology such as ATO, two approaches are possible – ‘local’, involving elements of a network such as trains and carriages, or migrating an entire network to a new status.”

Clearly, converting an entire network and its rolling stock to ATO will require time and patience before reaping any benefits. And the question here is: can the railways afford the wait?

Another factor to bear in mind is the 20 to 30-year lifetime of rail infrastructure and its average renewal rate of two to three per cent annually. This raises the question of duration of migration versus the lifetime of the technology.


One of the three functionalities of an automatic traffic management system, ATO concerns the non-safety part of train operation related to station stops and starts. In parallel, Automatic Train Supervision (ATS) is responsible for monitoring and controlling the rail system, and Automatic Train Protection (ATP) ensures trains remain at safe intervals and receive sufficient warning to allow them to halt without colliding.

At the Brussels meeting, participants debated the pros and cons of using ETCS (the European Train Control System installed on many main lines in Europe and beyond) as the ATP platform for main line rail ATO.

Originally designed with ‘trains plus drivers’ in mind, the standard version of ETCS, with its system requirements, architecture and interfaces, would have to be reviewed to specify the level of changes needed for ‘trains minus drivers’. Such a long regulatory process will obviously result in delays in bringing driverless rail into commercial service and obtaining any returns.

Remember, the reason for rail automation is, after all, to improve the competitiveness of the sector. “Ideally, we need a European system with a European system architecture so that locomotives can run seamlessly across borders, yet are uncoupled as far as possible from ETCS to avoid making migration more difficult than it already is,” pointed out Mr Doppelbauer.

Vladimir Kampik, director of European affairs for Czech operator AŽD Praha, described the use of ATO over ETCS as “the ideal partnership for safe, effective, precise, and comfortable train operations”.

AŽD Praha has two ATO systems in place, GTN (graphical and technological overlay) and AVV (Automatické Vedeni Vladu – automatic train control), covering 2,450km/25 per cent of the country’s rail network and carrying just over 70 per cent of rail traffic nationwide. 75 per cent of the company’s new and refurbished trains, 10 per cent of the entire fleet, are ATO equipped.

Mr Kampik insisted that ATO is a mature technology “as proven in its everyday operation on the Czech railway network for more than 25 years”. Future improvements, he suggested, should include providing dynamic train information along lines, to enable automatic corrective action, and introducing new functions according to Shift2Rail such as ‘virtually coupled trains’ and ‘ETCS Level 4’.

On the matter of ATO with ETCS, one downside in AŽD Praha’s experience is that the ETCS braking curves are not fully optimised for energy consumption savings. Yet, overall, the company views ATO/ETCS interconnection as the foundation for further development and deployment of ETCS, especially beyond Level 3. “ERA, together with the rail supply industry, operators and infrastructure managers, should ensure the interoperable use of ATO with freedom of choice of technology and implementation,” recommended Mr Kampik.

Integration & working together

French Railways (SNCF) runs automatic metro systems in France (Lyon, Lille and Rennes), through subsidiary Keolis, and is currently involved in projects in Shanghai and Hyderabad (India). In the field of driverless main line rail, it is gaining experience in the UK as part of the Thameslink franchise (35% Keolis, 65% Go-Ahead), and in France with the planned introduction of NExTEO technology automation (by Siemens) to the core of RER commuter Line E running through Paris.

Pierre Izard, chief technical
officer of SNCF, identified four key implementation issues and criteria for automatic main line rail, namely:

  • Conditions for traffic recovery after an incident and crisis management;
  • Reliability for both trains and infrastructure;
  • Legal and societal aspects;
  • Cyber security.

Yet, despite the above, SNCF is in favour of action to extend automation further afield. “ATO is not a dream but a major project. Train automation should be integrated into the European railway roadmap,” said Mr Izard. “In order to succeed, we need collaboration between all the European stakeholders and the strong involvement of ERA.”Human, social and environmental factors

One of the expected advantages of driverless operations is, of course, reduced staff cost. Yet automation is only relevant when it proves more effective than the human factor. In this case, the economics of such operation must be evaluated, scope- by-scope. Likewise when adapting driver staff volumes to traffic.

On this point, Laurent Dauby, rail director of the International Association of Public Transport (UITP), recommended entering into dialogue with unions and workers as early as possible in the process to discuss and explain ‘what’s in it for them’. Mr Izard added: “The goal of introducing ATO should not be to remove the human presence from the railways, although this may occur over time.” In any case, the human element will still be required for monitoring and exercising overall control.

On the environmental side, ATO is good news thanks to the optimisation of energy consumption, such as recovering braking energy that can subsequently be used for accelerating other trains. In this respect, AŽD Praha has some positive data on its fleet of automatic EMUs. “Our fleet of 83 Class 471 units runs 13 million kilometres and saves 14,000 tonnes of CO2 annually,” reported Mr Kampik.

Automation – the bigger and future picture

Mr Doppelbauer warned that automation in other sectors poses a real threat to the railways, particularly given the significant amount of R&D effort being poured into autonomous cars/trucks. Indeed, because of the current cost of the technology – a laser sensor for a Google car costs as much as US$75,000 (£52,000) – autonomous trucks rather than cars could hit the market first. Savings with autonomous trucks and truck convoying are expected to be greater than those to be achieved by rail with ATO.

In addition, Mr Izard warned that the balance of workforce savings and driverless operations is forecast to swing in favour of the roads (see box).

Workforce savings and driverless operations
Rail: 1 driver – 1,268 passengers Coach: 1 driver – up to 74 passengers
Car sharing: 1 driver – 4 passengers
Rail freight: 1 driver – payload up to 3,000 tonnes
Road haulage: 1 driver – payload of 25 tonnes
(Source: SNCF initial position paper, March 2016)

On a more positive note, since the train is already a guided mode of transport, there is no reason in theory preventing rail automation from developing as quickly as it is doing for the roads. “The railway system needs to be open to evolution, and automation is part of it,” summed up Michael Cramer, chairman of the RFE.

“The major benefits of ATO, such as capacity gains and energy savings, would definitely contribute to fulfilling the objectives set by the [European Commission’s] Transport White Paper.”