Much has been written and continues to be written on European Rail Traffic Management System (ERTMS) / European Train Control System (ETCS) and the benefits it can bring to the railway. In short these may be summarised as the reduction or even elimination of lineside signal infrastructure, the creation of greater capacity, and lower costs. All are admirable objectives but have the practicalities of achieving them really been thought through?
A major challenge could be optimising train timetable planning in an ETCS world. A recent talk given to the IRSE London & SE Section by Nadia Hoodbhoy, a principal engineer for Control & Command within Network Rail, was indicative of the multitude of factors needing to be considered.
ERTMS and the challenge
Just a brief reminder as to what ERTMS is all about. It has three component parts:
ETCS which is the signalling aspect and has three levels – L1, L2, and L3. L1 is essentially a standardised Train Protection System which is discontinuous but prevents drivers ignoring lineside signals. L2 provides Movement Authorities (MA) to drivers in the cab with fixed block sections and can result in lineside signals being removed or in conjunction with lineside signals. L3 moves toward a total radio-based system incorporating moving block where trains can close up in dense traffic areas and hopefully traditional train detection systems, e.g. track circuits and axle counters, can be dispensed with.
GSM-R and its intended replacement FRMCS, which is the radio link to connect all the data and transmit it between control centres and trains.
The European Traffic Management Layer (ETML) aimed at producing a standardised Traffic Management System (TMS) but which has never got off the ground. TMS information is alternatively provided by proprietary systems.
Pragmatism has introduced a variation on ETCS L2, known as Hybrid Level 3 in which a mixture of trains fitted either with L2 or L3 can operate on the same sections of track with an L3-fitted train followed by another L3 fitted train being given an MA that allows a closer spacing using moving block principles. Should the following train only be fitted with L2 equipment, then an MA can only be given to the end of a fixed block section. If all trains on the route are fitted with either L2 or L3, then lineside signals can be removed. However, it is more likely that the route will also carry non-fitted trains and thus signals have to be retained.
Widespread opinion is that conventional signalling is too expensive and that moving to ERTMS is the only way to significantly reduce costs. How is this measured? The method is by assessing the cost of a Signalling Equivalent Unit (SEU) that gives comparisons for different technologies to signal and control a typical section of railway. At present, an SEU costs in excess of £419,000 for conventional signalling (at pre-Covid price levels) and £315,000 for an ETCS Level 2 with signals removed. Even this is viewed as too much and a programme named Target 190 was launched in 2019 to reduce the SEU to £190,000 as part of the predicted future budget and in line with market evaluations at that time.
As part of this, there are three components:
- Technology embracing EULYNX as a standardised specification for interfaces and elements within railway signalling systems including interlockings, automatic train operation (ATO), traffic management systems, and optimised train detection methods.
- Data comprising signalling option selection and future control, command and signalling (F-CCS) timetabling.
- Process and tools for how F-CCS design and validation is decided together with the creation of option selection rules.
The Target 190 covers only the infrastructure elements of ETCS. The fitting of rolling stock has also to be considered and these are a significant cost but cannot be measured on a route mileage basis. Retrofitting and software upgrades are proving to be very expensive and only when a modular approach to fleet fitment is available will these reduce. Once a critical mass of trains are equipped, then the whole ETCS programme will show the benefits to be gained.
The aim is for ‘safe, affordable, and deliverable signalling’ combining ‘innovation, standardisation, and automation’. Some may view this as patently obvious and in essence it has always been part of a forward-looking signalling concept. So, what is different now and how (maybe if) it will be achieved?
Engaging the industry
The target is to come up with a generic operational concept that will feed into all future signalling projects and enable business continuity. Delivery support will be crucial and thus signalling engineers from both the infrastructure owner and supply chain industries will lead to a train control system framework.
The principles for network change are already in place but are rarely co-ordinated so these need to be linked to long-term deployments. As has been stated many times in the past, the collection and accuracy of data is essential for asset information and improvements to this task are already underway in a series of projects to be delivered through the Network model.
At the current time, all of this is conceptual but would lead to scheme re-design options with continual feedback and adjustment mechanisms in place. A market study into timetabling and train planning tools took place in 2024 which resulted in a requirement for new methodology and tools being required. From this, the development of an integrated tooling ecosystem to underpin the end-to-end design and validation process is underway and will be known as the Synthetic Environment.
Horizon scanning and development of the proof of concept for this has been completed. The principal suppliers were consulted by way of interviews and questionnaires with observers present. New assistive tools, specialised train planning tools and microscopic operation simulation tools would be required, all leading to a revised way as to how re-signalling schemes are progressed.
Timetable planning
If the increased capacity claim that ERTMS promises is to be realised, then the timetable planning process must also change. The existing tools for timetable compilation are designed for conventional signalling and need to be modified for ETCS operation. Different driver behaviour will emerge and much greater opportunities for ATO should be part of the deliverables. Currently, there is poor integration with the scheme design process.
How then to modernise train operational planning to unlock the full benefits. The work to achieve this is viewed as a three-stage approach:
- Specify high level targets for F-CCS and from this produce timetable recommendations.
- From these recommendations, work out the needs of F-CCS and produce a timetabling road map.
- Undertake a market study to explore how a timetable partner evaluation can be achieved.
These factors are being studied by Birmingham and Delft universities in collaboration with Network Rail and other international rail communities. Feedback from operational experience on the Cambrian Line and the Thameslink central core (both with several years of ETCS operation but very different kinds of railway – one a rural backwater, the other a high-density urban railway) has proved useful.
The timetabling developments seen as needed must take account of:
- In-cab signalling and resultant higher speeds.
- Full supervision of train movements meaning that trains can run closer together and more flexible use of the track.
- Automated operational control which with ATO leads to greater precision as to a train’s whereabouts.
Three principles then emerge:
A much closer alignment of signal engineers and operators will be needed and, importantly, operators will need to understand what ERTMS can offer. The relationship between systems that provide ETCS technology, traffic management (TMS), driver advisory systems (DAS), train path allocation, rolling stock characteristics, braking performance, routing, platforming, and run times have to be harmonised or even incorporated into a single system.
KPIs would be required and set down during the strategic planning stage. These would lead to different upgrade options, each of which would need to be assessed in line with operators submitting train path requests.
Currently, no such single system exists that is capable of covering these requirements and it will require a collaborative development including industry and innovation partners if one is to emerge. The challenge is considerable as there are so many factors and variables in the requirements list.
Change required
For the UK, the transition to Great British Railways (GBR) will assist in focussing minds but considerable business change will also be needed. It is interesting that in Germany a similar Capacity and Traffic Management system to automate planning and dispatching requirements was put out to industry with the result that no such product existed and thus no offer was forthcoming.
This demonstrates the challenge for industry so, if this article spurs any organisation out there to believe they can offer a solution, I am sure Network Rail (and GBR) would be delighted to listen!
Thanks are extended to Nadia for supplying additional facts and ensuring accuracy of content.
Editor’s note: This article reflects Network Rail’s view as specified in its ‘CP7 delivery plan – Our approach to digital signalling’. However, rolling stock sources and the ‘Signalling Scotland’s Future’ report produced by Network Rail Scotland have a different perspective, particularly in respect of affordability. This difference is explained in our feature “Is ETCS affordable?”
Image credit: iStockphoto.com/BIM