GoA 2 trains have been operating in Europe for a number of years.
GoA 2 trains have been operating in Europe for a number of years.
( Source: gemeinfrei / Pexels)

AUTONOMOUS TRAINS Driverless trains in Europe: heavy rail automation explained

Author / Editor: Jamie Thomson / Isabell Page

Heavy rail automation has been a widely discussed topic in recent years. As an increasing number of light rail networks become autonomous, it’s only a matter of time before the heavy rail sector follows suit. This article explores the current state of heavy rail automation in Europe.

Like self-driving cars, there are five levels of Automatic Train Operation (ATO). Grade of Automation (GoA) 0, is comparable to an electric tram running through city streets, where the train is operated by a driver on-site. GoA 1 is where a train is operated manually. The driver is responsible for starting and stopping the train and operating doors etc.

GoA 2 refers to semi-automatic operation, where starting and stopping are automated, but the driver is responsible for operating doors and handling emergencies. GoA 3 is where trains are driverless. Starting and stopping is automated and an attendant operates the doors. GoA 4 is the highest level of train automation where the train is operated unattended. All operations are automated, including in emergencies.

According to a recent report published by the Japan Transport and Tourism Research Institute (JTTRI), There are currently 64 fully automated train lines (both light and heavy rail) in the world, with more than half located in Asia.

How does heavy rail automation work?

Automated train lines need to be connected to a Traffic Management System (TMS) that can monitor train journeys and detect obstacles and conflicts. The system needs to be able to resolve conflicts as they happen in order maintain schedules and keep services running smoothly.

Railway traffic systems require a large amount of interactions, or touch points, in order to generate data on the status of journeys and the trains themselves. Artificial Intelligence (AI) plays an essential role in ensuring that trains communicate effectively with sensors and provide feedback. Through Machine Learning (ML), big data analysis and pattern recognition, European Train Control Systems (ETCS) are able to carry out heuristics and simulations to optimize operations.

AI will continue to play a crucial role in heavy rail automation in the future, so much so, that Europe’s rail research company Shift2Rail has recently launched the RAILS project to further explore the potential of AI in European rail travel.

The benefits and challenges of heavy rail automation

Fully-automated trains can improve the capacity of rail systems and make services more reliable. For example, if the system detects and issue with a particular service, additional trains can be deployed to ensure schedules continue to run as planned.

Heavy rail automation can also be more energy efficient, improve safety and be more cost effective in the long run. One of the biggest challenges of heavy rail automation is implementing the necessary infrastructure. Automated trains operate in a more complex environment than mainline railways and as such, require different train types, stopping distances and of course, more technology, all of which come at a price.

Safety is of paramount importance in ATOs, and rail lines need to be physically distanced from the surrounding environment. For example, existing level crossings and bridges may need to be rebuilt in order to adhere to spatial guidelines.

Heavy rail automation in Europe

Whereas driverless trains are currently in operation in light rail networks in Germany, France and Italy, heavy rail systems have yet to be made fully autonomous. However, in the Czech Republic, heavy rail networks have been operating GoA 2 semi-automatic trains since 1991 and in the UK, rail operator Thameslink has been operating GoA 2 since 2019, with plans to upgrade to GoA 4 in the future. The Netherlands is also currently running GoA 2 trials for both passenger and freight trains.

In France, the French National Railway Company (SNCF) recently completed a test run of a remotely controlled autonomous train. The trial was conducted as part of an initiative to develop GoA 4 prototypes by 2023. The locomotive-hauled train travelled a distance of 4km using satellite and 4G technology:

In Germany, German Rail (DB) is currently working with technology company Siemens, to develop a fully-automated S-Bahn line in Hamburg by 2021. Initially, it is planned that four trains will travel 500 m between the siding and platform at the intermediate station at Bergdorf, without any attendants on-board. Following a successful pilot phase, German Rail plans to automate the entire Hamburg S-Bahn network.

In summary

In some European countries, semi-autonomous trains have been in operation for several years. In recent times, an increasing number of countries are carrying out GoA 2 trials with plans to integrate semi-autonomous trains into their mainline networks. Although there are no GoA 4 trains currently operating in Europe, prototypes are being developed and we may see fully-automated trials being carried out as early as 2023.