ERTMS/ATO is the European interoperable solution for automatic train operation. It is designed to work together with ERTMS/ETCS, which provides the train protection layer and transmits the signalling constraints that automatic driving must respect.
However, ERTMS/ATO depends on the availability of ERTMS/ETCS. In practice, this creates a migration issue. Many lines are still operated with national signalling systems and lineside signals, while ERTMS/ETCS trackside deployment takes time and requires significant investment.
This raises a simple question: How can railway operators start using ERTMS/ATO before the full deployment of ERTMS/ETCS trackside?
Lineside Signalling Interpretation is one possible answer. It is not a replacement for ERTMS/ATO, nor a competing automation system. It is a migration concept aiming to reuse the standard ERTMS/ATO solution on lines that are not yet equipped with ERTMS/ETCS trackside, by interpreting existing lineside signals and converting their information into data usable by an ERTMS/ETCS onboard unit.
In this article, we will explore why this migration concept is needed, what Lineside Signalling Interpretation means, and how it could support the transition towards interoperable railway automation.
If you are not familiar with ERTMS/ATO, I recommend reading first:
- ERTMS: the European Rail Traffic Management System
- ERTMS/ETCS: the European Train Control System
- ERTMS/ATO: Europe’s Interoperable Train Autopilot
Article summary
ERTMS/ATO is the European standard solution for automatic train operation. It relies on ERTMS/ETCS to provide the train protection layer and the signalling information required by the ATO system.
This architecture is coherent for the target European railway system. However, its deployment depends on the availability of ERTMS/ETCS trackside. As long as large parts of the network remain operated with legacy signalling systems, the use of standard ERTMS/ATO remains limited.
Several migration solutions have therefore been investigated in Europe. Their objective is to enable automatic driving on lines not yet equipped with ERTMS/ETCS, while reusing as much as possible the standard ERTMS/ATO architecture and interfaces.
Lineside Signalling Interpretation follows this logic.
The core idea is to equip the train with an ERTMS/ETCS onboard unit, and to feed it with signalling information derived from existing lineside signals. A perception component detects the aspect displayed by the signal. A signalling converter then translates this information into ERTMS/ETCS-compliant data, using the train localisation and infrastructure data stored in a Digital Map.
From the point of view of the ERTMS/ETCS onboard unit, this concept can be seen as an onboard emulation of trackside ERTMS/ETCS Level 1. The traditional trackside chain — signal, LEU and Eurobalise — is replaced by a train-centric chain involving perception, localisation, Digital Map and signalling conversion.
This migration concept offers several potential benefits:
- it reuses the standard ERTMS/ATO solution;
- it reduces dependency on immediate ERTMS/ETCS trackside deployment;
- it supports a progressive rollout of onboard ERTMS/ETCS;
- it prepares future migration towards radio-based ERTMS/ETCS and higher Grades of Automation.
However, it is not the target architecture for high-capacity operation. It cannot provide the same headway performance as radio-based ERTMS/ETCS levels. It also requires further work on operational rules, hazard analysis, interfaces, perception confidence and certification.
Lineside Signalling Interpretation should therefore be understood as a migration concept: a way to accelerate the transition towards interoperable ERTMS/ATO, not as a substitute for the future European railway target system.
Last modified: 2026-05
This article by Bastian Simoni is licensed under CC BY-NC-SA 4.0
Content
- Why: the migration problem
- ERTMS/ATO as the European standard solution
- The dependency on ERTMS/ETCS
- ERTMS/ETCS trackside rollout takes time
- What: Lineside Signalling Interpretation
- The basic idea
- What the concept is not
- A train-centric approach
- How: the system concept
- Main building blocks
- ERTMS/ATO on-board
- PER: the perception component
- SCV: the signalling converter
- Digital Map
- Localisation
- Operational Execution
- From lineside signal to ERTMS/ETCS data
- The functional chain
- Comparisation with ERTMS/ETCS Level 1
- Virtual ETCS area
- Migration benefits
- Reusing the standard ERTMS/ATO solution
- Reducing dependency on trackside deployment
- Encouraging on-board ERTMS/ETCS rollout
- Supporting the transition towards GoA3 and GoA4
- Lessons from previous experiments
- The Dutch experiment: ATO over ATB
- The French autonomous train projects
- Limitations and open points
- Not the same capacity as radio-based ERTMS/ETCS
- Limited to areas equipped with light signals
- Data quality and application engineering
- Perception and confidence
- Operational rules and hazards analysis
- System Architecture perspective
- Conclusion
1. Why: the migration problem
1.1 ERTMS/ATO as the European standard solution
The standard ERTMS/ATO solution was designed to add automatic train operation to the ERTMS ecosystem.
Its objective is to provide an interoperable and interchangeable solution for mainline railway automation. Instead of developing national or proprietary automatic driving systems, the European approach is to define a standard solution working together with ERTMS/ETCS.
This is essential for the national railway network.
Unlike metro systems, mainline railways are open and heterogeneous environments. They involve multiple operators, different types of rolling stock, mixed traffic, cross-border services and long migration periods.
In this context, automatic train operation cannot be treated as an isolated onboard function. It must be interoperable with the signalling system, the traffic management system, the infrastructure data and the operational rules.
This is the purpose of ERTMS/ATO.
1.2 The dependency on ERTMS/ETCS
ERTMS/ATO relies on ERTMS/ETCS to provide the safety supervision layer.
ATO drives the train. ETCS supervises the train.
This distinction is fundamental. ATO manages traction and braking in order to follow the mission, respect timing points and optimise the driving profile. ETCS supervises the movement authority, speed limits and braking curves. If ATO does not respect the permitted envelope, ETCS intervenes.
This means that standard ERTMS/ATO requires an ERTMS/ETCS onboard unit. It also requires the ERTMS/ETCS conditions allowing automatic driving to be fulfilled.
On lines equipped with ERTMS/ETCS, this architecture is coherent.
But on lines not yet equipped with ERTMS/ETCS trackside, the situation becomes more difficult. The train may have an ATO system, but the signalling information required for safe automatic driving is not available through the standard ERTMS/ETCS chain.
This creates a deployment barrier.
1.3 ERTMS/ETCS trackside rollout takes time
ERTMS/ETCS has been designed to replace national signalling systems and enable interoperability across Europe. However, trackside deployment is a long and costly process.
In practice, national systems will remain in operation for many years.
This creates a paradox.
On the one hand, ERTMS/ATO is the target interoperable solution for automatic train operation. On the other hand, its use is constrained by the pace of ERTMS/ETCS trackside deployment.
For railway undertakings, this can delay the benefits of automatic driving:
- reduced energy consumption;
- better timetable adherence;
- preparation for more automated railway operations.
This is why migration solutions are needed.
The objective is not to bypass ERTMS/ATO, but to make it usable earlier, while still converging towards the European target system.
2. What: Lineside Signalling Interpretation
2.1 The basic idea
Lineside Signalling Interpretation is a migration concept enabling ERTMS/ATO on lines that are not yet equipped with ERTMS/ETCS trackside.
The idea is to use existing lineside signals as an input to an onboard system.
Instead of receiving ERTMS/ETCS data from trackside Eurobalises or a Radio Block Centre, the train interprets the aspect displayed by the lineside signal. This information is then converted into ERTMS/ETCS-compliant data and transmitted to the ERTMS/ETCS onboard unit.
The ERTMS/ETCS onboard unit can then provide the information required by ATO through the standard interface.
In simple terms:
- the signal displays an aspect;
- the train detects this aspect;
- the system interprets it;
- the aspect is converted into ERTMS/ETCS information;
- ETCS supervises the train;
- ERTMS/ATO can drive the train.
The key point is that ERTMS/ATO remains the standard solution. The migration concept does not change ATO itself. It changes the way the ERTMS/ETCS onboard unit receives the signalling information it needs.
2.2 What the concept is not
It is important to clarify what Lineside Signalling Interpretation is not.
It is not a new proprietary automatic driving system.
It is not a solution that revitalises class B systems.
It is not an alternative to ERTMS/ATO.
It is not the final high-capacity architecture for the future European railway network.
It is a migration concept.
Its purpose is to support the use of ERTMS/ATO before the full deployment of ERTMS/ETCS trackside, while keeping the target architecture as standard and interoperable as possible.
This is why the concept is particularly interesting from a system architecture perspective. It does not try to create another automation stack. Instead, it tries to reuse the ERTMS/ATO stack by adding a migration layer between existing lineside signalling and the ERTMS/ETCS onboard unit.
2.3 A train-centric approach
In traditional ERTMS/ETCS Level 1, the trackside system transmits information to the train through Eurobalises.
For each signal, a Lineside Electronic Unit reads the signal aspect and selects the appropriate ERTMS/ETCS telegram. This telegram is then transmitted to the train through Eurobalises.
The application engineering process defines which ERTMS/ETCS data correspond to each possible signal aspect.
Lineside Signalling Interpretation follows a similar logic, but moves part of the acquisition and interpretation chain onboard.
Instead of relying on a trackside LEU and Eurobalises for each signal, the train uses onboard perception, localisation, Digital Map data and a signalling converter.
This is why the concept can be understood as a train-centric emulation of ERTMS/ETCS Level 1.
It does not remove the need for engineering. On the contrary, application engineering remains essential. The system must still know which ERTMS/ETCS data are associated with each signal aspect and each location.
But the way this information reaches the ERTMS/ETCS onboard unit changes.
3. How: the system concept
3.1 Main building blocks
The concept relies on several main building blocks.
ERTMS/ETCS onboard
The train is equipped with an ERTMS/ETCS onboard unit.
This is a major difference compared with some early migration experiments based on an adapter replacing the ETCS onboard function from the ATO point of view.
In the Lineside Signalling Interpretation concept, the onboard ATP remains ERTMS/ETCS. This is important because it supports convergence towards the target European system.
ERTMS/ATO onboard
The ATO onboard system remains the standard ERTMS/ATO solution.
It receives signalling information from the ERTMS/ETCS onboard unit through the standard ATP/ATO interface. From the ATO point of view, the objective is to avoid creating a specific migration version of the ATO system.
PER: the perception component
PER is the perception component.
Its role is to acquire the aspect of the target lineside signal from the physical environment.
In other words, PER detects what the signal is displaying.
This may involve computer vision and perception algorithms, but the article should not reduce the concept to vision technology. PER is only one part of the system. The key architectural challenge is not only to see a signal, but to determine which signal is relevant, interpret its aspect and convert the result into usable ERTMS/ETCS information.
SCV: the signalling converter
SCV is the signalling converter.
Its role is to translate the signal aspect information into ERTMS/ETCS-compliant data.
SCV uses several inputs:
- the train localisation;
- infrastructure data from the Digital Map;
- the signal aspect acquired by PER;
- and, where available, expected signal aspect information from operational systems.
Based on this information, SCV selects the ERTMS/ETCS data corresponding to the detected signal aspect and sends them to the ERTMS/ETCS onboard unit.
SCV is therefore the architectural bridge between legacy lineside signalling and the ERTMS/ETCS onboard unit.
Digital Map
The Digital Map provides infrastructure data.
It contains the information required by the onboard system to understand the railway environment: track geometry, signal locations, signalling information and other data required for the interpretation process.
The Digital Map is not only a navigation support. In this concept, it is a key element of the signalling interpretation chain.
Without accurate infrastructure data, the system cannot reliably associate a perceived signal with the correct track, the correct route and the correct ERTMS/ETCS information.
Localisation
Localisation provides the position and heading of the train.
This is essential because the onboard system must know where the train is in relation to the infrastructure and to the target signal.
Perceiving a signal is not sufficient. The system must also know whether this signal is the relevant one for the train, and what ERTMS/ETCS information must be associated with it.
Localisation, perception and Digital Map data therefore work together.
Operational Execution (Traffic Management System gateway)
Operational Execution is connected to the traffic management environment.
It provides the Journey Profile, which defines the route to be followed. Depending on the implementation, it may also provide expected signal aspect information.
This information can be used to support the interpretation process and to check consistency between what is expected and what is perceived.
4. From lineside signal to ERTMS/ETCS data
4.1 The functional chain
The functional chain can be described as follows.
First, the train receives its mission and route information through the Journey Profile.
Then, the onboard system uses this Journey Profile to retrieve the relevant infrastructure data from the Digital Map.
The train localises itself on the route.
Based on the route, the localisation and the Digital Map, the system determines the next target signal.
PER searches for this signal in the physical environment and reads its displayed aspect.
SCV receives this perceived signal aspect, checks it against the available data, and selects the corresponding ERTMS/ETCS information.
The ERTMS/ETCS onboard unit receives this information and can supervise the train accordingly.
Finally, ERTMS/ATO can use the information provided by ETCS to perform automatic driving.
This chain is the core of Lineside Signalling Interpretation.
4.2 Comparison with ERTMS/ETCS Level 1
The concept is easier to understand when compared with ERTMS/ETCS Level 1.
In Level 1, the signal aspect is acquired trackside. A Lineside Electronic Unit is connected to the signal and selects the ERTMS/ETCS information corresponding to the current aspect. This information is then sent to the train through Eurobalises.
In Lineside Signalling Interpretation, the signal aspect is acquired onboard. The perception system reads the signal. The signalling converter selects the corresponding ERTMS/ETCS information using localisation and Digital Map data. This information is then sent to the ERTMS/ETCS onboard unit.
The philosophy is similar:
- interpret the signal aspect;
- associate it with ERTMS/ETCS data;
- provide this data to the onboard ETCS.
The implementation is different:
- in Level 1, the acquisition chain is mainly trackside;
- in Lineside Signalling Interpretation, the acquisition chain becomes train-centric.
This is why the concept can be described as a virtualised approach to ETCS Level 1.
4.3 Virtual ETCS Area
A useful concept is the Virtual ETCS Area.
A Virtual ETCS Area is an area where the onboard system emulates the role of ERTMS/ETCS trackside equipment.
There can be two main cases.
The first case is a Virtual ETCS Area without signal detection. In this case, SCV can provide ERTMS/ETCS data contained in the Digital Map, based on the localisation of the train.
The second case is a Virtual ETCS Area with signal detection. In this case, SCV provides ERTMS/ETCS data based on both the localisation of the train and the aspect of the target light signal acquired by PER.
The second case is the most relevant for lines equipped with lineside signals.
A Virtual ETCS Area with signal detection can be understood as an onboard emulation of ERTMS/ETCS Level 1 trackside equipment. The physical Eurobalises and LEU are replaced by a combination of localisation, Digital Map, perception and signalling conversion.
This does not mean that all trackside equipment disappears. Some physical Eurobalises may still be required, for example for relocation purposes. But the concept changes the way signalling information is acquired and transmitted to the ERTMS/ETCS onboard unit.
5. Migration benefits
5.1 Reusing the standard ERTMS/ATO solution
The first benefit is the reuse of ERTMS/ATO.
The objective is not to develop a specific ATO system for each national signalling system. Such an approach would reproduce the fragmentation that ERTMS is precisely intended to overcome.
By feeding an ERTMS/ETCS onboard unit with information derived from lineside signalling, the system allows ERTMS/ATO to remain standard.
This is important for interoperability and long-term maintainability.
The migration layer may be specific to the transition period, but the ATO solution remains aligned with the European target architecture.
5.2 Reducing dependency on trackside deployment
The second benefit is the reduction of dependency on immediate ERTMS/ETCS trackside deployment.
ERTMS/ETCS trackside rollout will continue. It remains the target direction for the European railway system.
However, railway undertakings may want to obtain the benefits of automatic driving before the full completion of this rollout.
Lineside Signalling Interpretation offers a possible intermediate step.
It allows trains equipped with ERTMS/ETCS onboard and ERTMS/ATO to use existing lineside signalling as a source of signalling information, through perception and conversion.
This can support a progressive deployment strategy.
5.3 Encouraging onboard ERTMS/ETCS rollout
This concept could also support the rollout of ERTMS/ETCS onboard equipment.
If railway undertakings can obtain operational benefits from equipping trains with ERTMS/ETCS onboard before the full deployment of ERTMS/ETCS trackside, the business case for onboard equipment becomes stronger.
The train would already be prepared for the future target architecture:
- ERTMS/ETCS onboard;
- ERTMS/ATO onboard;
- localisation;
- perception;
- radio communication;
- Digital Map-based operation.
This could make the later transition towards radio-based ERTMS/ETCS and higher Grades of Automation easier.
5.4 Supporting the transition towards GoA3 and GoA4
The concept is especially relevant when considering the long-term transition towards GoA3 and GoA4.
In GoA2, the driver remains in the cab and supervises the operation. The system automates traction and braking.
In GoA3 and GoA4, the system must support a much higher level of automation. This requires new capabilities, including perception, localisation, hazard detection, operational decision-making and integration with traffic management.
Lineside Signalling Interpretation does not solve all these challenges. But it contributes to the migration path by introducing some of the building blocks that will be required for more automated operation.
It is therefore not only a short-term solution for GoA2. It is also a step towards the broader system architecture required for future autonomous railway operation.
6. Lessons from previous experiments
6.1 The Dutch experiment: ATO over ATB
The Dutch railway undertaking NS experimented with an adapted ATO over ETCS solution between 2020 and 2021.
The objective was to use the standard ATO over ETCS approach on a train and a track not equipped with ERTMS/ETCS.
In this experiment, the absence of ETCS onboard was compensated by an adapter. This adapter collected signalling information from several sources and converted it into information compatible with the ATO interface.
The system used information from:
- lineside signalling;
- the national ATB system;
- interlocking and block occupancy systems.
This experiment showed that it is possible to reuse the ATO over ETCS principles in a migration context. It also highlighted important questions: dependency on national systems, data availability, human factors and the long-term transition towards the standard ERTMS target system.
6.2 The French autonomous train projects
In France, automatic driving experiments have also investigated the use of lineside signalling perception and interpretation.
The Autonomous Freight Train project demonstrated automatic driving using an adapted ATO over ETCS solution over French lineside signalling.
These experiments are important because they show the practical interest of migration solutions. They also illustrate the need to move from project-specific adaptations towards more standardised concepts.
This is where the Lineside Signalling Interpretation concept becomes important.
It provides a more structured system approach: instead of only adapting ATO to existing systems, it investigates how to feed an ERTMS/ETCS onboard unit with information derived from lineside signalling.
7. Limitations and open points
7.1 Not the same capacity as radio-based ERTMS/ETCS
Lineside Signalling Interpretation is a migration concept. It should not be confused with the final high-capacity architecture.
Because it emulates a Level 1-like approach, it cannot achieve the same headway performance as radio-based ERTMS/ETCS levels.
For maximum line capacity, ERTMS/ETCS Level 2 and future radio-based evolutions remain the target.
Lineside Signalling Interpretation can help start automatic driving earlier. It can support migration. But it is not the final answer for high-density railway corridors where continuous radio-based supervision is required.
7.2 Limited to areas equipped with light signals
The concept applies to areas equipped with lineside light signals.
Areas without light signals, for example where cab signalling is fully ensured by a national Class B system, are outside the concept scope.
This is an important limitation.
The system needs something to perceive. If there is no lineside signal to interpret, another migration approach is required.
7.3 Data quality and application engineering
The concept relies heavily on infrastructure data.
Signal locations, signal characteristics, track geometry, route information and ERTMS/ETCS application data must be accurate and maintained.
This is not entirely new. ERTMS/ETCS Level 1 deployment already requires a strong application engineering process. But in a train-centric approach, the quality and availability of Digital Map data become even more central.
The migration concept therefore raises questions about:
- data ownership;
- data maintenance;
- configuration control;
- certification;
- updates over time;
- consistency between operational data and physical infrastructure.
7.4 Perception and confidence
Perception is a key enabler, but it also raises technical and safety challenges.
The system must detect the correct target signal, read its aspect, assess confidence, and manage discrepancies between the perceived aspect and the expected aspect.
This is not only a computer vision problem.
It is a system problem involving perception, localisation, Digital Map data, operational context and safety logic.
Further work is therefore needed on perception confidence, expected signal aspect confidence, target signal selection and the management of unexpected or temporary signalling.
7.5 Operational rules and hazard analysis
The concept still requires further work before it can become a complete specification.
Operational rules must be defined.
Hazard analysis must be performed.
Interfaces must be specified.
Failure modes must be analysed.
The relationship between safety, operability and degraded modes must be clarified.
This is normal for a migration concept at this level of maturity. The important point is not to hide these open questions, but to identify them clearly.
8. System architecture perspective
Lineside Signalling Interpretation is interesting because it is not only a technical trick to read signals with a camera.
It is an architecture concept.
It tries to solve a migration issue by inserting a new layer between existing signalling and the ERTMS/ETCS onboard unit, while preserving the standard ERTMS/ATO solution as much as possible.
This makes the concept a good example of railway system-of-systems thinking.
It involves:
- existing infrastructure;
- onboard train protection;
- automatic train operation;
- perception;
- localisation;
- Digital Map;
- traffic management;
- application engineering;
- safety and certification;
- migration strategy.
The difficulty is not only to make each component work. The difficulty is to make the whole system coherent, certifiable, interoperable and useful during a long transition period.
This is why Lineside Signalling Interpretation should be analysed less as a perception function, and more as a migration architecture towards ERTMS/ATO.
Conclusion
ERTMS/ATO is the European interoperable solution for automatic train operation. It is the target architecture for mainline railway automation, because it allows automatic driving to be integrated into the broader ERTMS ecosystem.
However, ERTMS/ATO depends on ERTMS/ETCS. As long as many lines are not equipped with ERTMS/ETCS trackside, the deployment of standard automatic train operation remains constrained.
Lineside Signalling Interpretation addresses this migration issue.
The concept consists in interpreting existing lineside signals, converting their aspects into ERTMS/ETCS-compliant information, and feeding an ERTMS/ETCS onboard unit. The ERTMS/ETCS onboard can then supervise the train and provide the information required by ERTMS/ATO.
This approach can be understood as a train-centric emulation of ERTMS/ETCS Level 1.
It offers a possible migration path:
- reuse the standard ERTMS/ATO solution;
- equip trains with ERTMS/ETCS onboard earlier;
- reduce dependency on immediate trackside deployment;
- prepare future migration towards radio-based ERTMS/ETCS;
- introduce building blocks needed for higher levels of automation.
But this concept is not the final target system.
It will not provide the same capacity performance as radio-based ERTMS/ETCS. It is limited to areas equipped with light signals. It requires high-quality infrastructure data, safe localisation, reliable perception, robust conversion logic, operational rules and hazard analysis.
Its value is therefore not to replace ERTMS/ETCS trackside deployment.
Its value is to support the transition.
Lineside Signalling Interpretation is a migration bridge between today’s lineside signalling and tomorrow’s interoperable automated railway system.
In that sense, it is one of the most interesting examples of how railway automation should be approached: not as a standalone technology, but as a system architecture problem.
Lineside Signalling Interpretation documentation
Documentation from European collaborative R&D programmes
Europe’s Rail Joint Undertaking programme, 2021+
Shift2Rail programme, 2014–2021
- TAURO project: technologies supporting the migration to ERTMS/ATO
- Lineside Signalling Interpretation report
- Digital Map report