Today’s railway technology can enable fully automated trains that travel at up to 350 km/h across very long distances. But such high speeds make it harder for drivers to respond to visual trackside signals. Higher levels of automation are therefore required for safety reasons. The emergence of long distance cross-border routes also creates problems for existing signalling systems.
“In Europe today, there are a lot of countries with different Class B train control systems which are not interoperable,” explains Olaf Mense, senior expert in train control systems at Siemens.
“If you want to run a train from, say, Genoa to Amsterdam, you need a multi-system locomotive with three or four different systems. They require different antennas under the train and different sensors, and there may not be enough space for the distance required between multiple antennas because of interference,” he says. “This is a problem for cross-border rail traffic.”
A train running from Genoa to Amsterdam would have to be compatible with legacy systems like SCMT (Sistema Controllo Marcia Treno) in Italy, ZUB-262 in Switzerland, PZB (Punktförmige Zugbeeinflussung) in Germany, and ATB (Automatische TreinBeïnvloeding) in the Netherlands. More than 20 such systems are in use in Europe today. Clearly, an interoperable automated signalling solution is required.
The proposed European Rail Traffic Management System (ERTMS) consists of two technologies. GSM-R, which has been widely rolled out, is an adaptation of standard GSM for use in rail systems. The less widely adopted European Train Control System (ETCS) transmits data to the train from radio beacons (balises) positioned along the track or via GSM-R.
An ETCS system like Siemens’ Trainguard 200 comprises both on-board and trackside electronics (Fig. 1). A European Vital Computer (EVC) is at the centre of the system, with a visual interface for the driver (Driver Machine Interface or DMI) in the cab (Fig. 2). Trackside electronics include the lineside electronic unit (LEU) and the ETCS-compatible balises, called Eurobalises.
Balise Operation
Eurobalises have a very important function in ERTMS: they tell the train exactly where it is and what’s coming up on the line ahead. Mounted in the centre of the track, each features an antenna and an encoder/decoder for transmitting telegrams (Fig. 3). The balise is powered by electromagnetic induction from the antenna on the underside of the train as it passes overhead at speeds of 250 km/h and beyond. It uses a modulated RF signal at 27 MHz.
Two types of balise exist. “Fixed” balises transmit the same telegram to every train, but a telegram from a “transparent” balise is variable and dictated by the LEU positioned next to the track. Mean time to failure (MTTF) for these small passive beacons is typically 20 years for transparent or 30 years for fixed balises, and they have to operate under extremely harsh conditions from snow to lightning to solar radiation. Rugged design, then, must be carefully considered.
“It’s difficult to design [for these conditions], but it’s normal for our business,” says Mense. “It’s not new that we have strong variability in the power supply, a wide range of temperature and vibration. It’s normal. There is a lot of experience in our company for this.”
At the moment, two levels of ETCS functionality are operating in Europe. Level 1 relies entirely on balises, and level 2 uses balises plus GSM-R. The telegrams used in both levels are identical and do not depend on the method of delivery.
“In level 1, the main task of the balise is to transmit information regarding the movement authority, so you get information about the track ahead, for example, where the next signal is,” explains Mense. A level 1 telegram might contain data like the location, the track gradients, speed limits, braking curves, and distance to the next set of points.
“In level 2, the main task of the balise is as an electronic location reference, a milestone, for the system. You get the movement authority via GSM-R, but there’s always a starting point from which this information is valid, and that’s always the balise,” he adds.
Positioning Subsystem
Since there can be several kilometres between the balises, locating a train accurately requires an additional odometry subsystem.
“The odometry subsystem calculates the distance travelled by the train. All the positioning of the train is done by this subsystem until it reaches a balise,” explains Victor Martinez, global business development manager for transport systems at mechatronic specialists SKF.
The odometry calculations rely on signals from position sensors, usually either radar (Doppler) sensors or wheel speed pulse sensors. “If these signals are not reliable, all the calculations done by the software will be wrong and the train can get confused,” Martinez adds.
Developed for the train axle box, SKF’s Axletronic mechatronic wheel pulse sensor can function as a speed sensor and an odometer in ETCS systems. The double channel sensors are based on magnetic technology and incorporate two transducers on a small printed-circuit board (PCB).
When the wheel rotates, each transducer receives magnetic pulses, with a slight time shift (Fig. 4). The odometry subsystem can tell which direction the train is travelling in, depending on which transducer peaks first. Any slight inaccuracies in distance are corrected when the train is “reset” to a known position when it passes a balise.
According to Martinez, magnetic technology offers several advantages over optical sensors.
“The main one is that there is no wear inside the sensor so it’s maintenance free,” he says. “This means that the MTTF is very long. You can also put a magnetic sensor in the same body as a temperature sensor, which for high-speed trains is a must.”
Isolated Installations
Although GSM-R has been widely adopted, ETCS so far only has been rolled out to a small proportion of European track, around 4400 km. For example, the Madrid-Barcelona line in Spain and the Rome-Naples line in Italy use ETCS level 2. But these are isolated installations.
This is partly because the technology is only going into new trains. There’s a lack of strong business reasons for train operators to invest in upgrading trains and track. Retrofitting an existing locomotive is not easy, since additional space must be found, without removing any existing train control systems. It’s much easier to include the technology in the design phase for a new train.
Difficulties with upgrading trains are coupled with a complex and time-consuming approvals process for cross-border locomotives.
“You have to go to the national safety authorities [in each country], and they each have special requirements for the approval. It’s not harmonised,” says Siemens’ rail expert Olaf Mense. “It’s a big problem at the moment since there is no cross-acceptance between the countries.”
So, how long will it take before we have truly interoperable train systems in Europe? Waiting for equipment on all the main rail routes in Europe to reach its end-of-life may take a really long time. As an example, an ETCS level 2-compatible system will replace Denmark’s entire rail signalling network, which is obsolete. This process began in 2008 and will take until 2021. A lack of commitment from some European countries that already have well-developed train control systems in place may mean it takes another 20 or 30 years, by which time technology probably will have moved on again.