IMPRESS - IMproving the Performance of the Railways through an EGNSS Safety Service

The objective of IMPRESS project was to assess the feasibility of an EGNSS-based safety service for the rail sector beyond 2022. This service would have the potential to improve the functionalities of the rail signaling infrastructure. 

Contract Number: Call for Tenders No 762/PP/GRO/19/11304

Project Segment: H2020 EGNSS Mission & Services

Duration: 35 months (September 2020 – July 2023)

Budget: €379 567

Project Partners: Egis Avia (France), Egis Rail (France), Thales Alenia Space (France), GMV NSL (Spain)

• Project Coordinator:
  Jonathan Vuillaume, Egis Avia (France)
   
• European Commission Project Manager:
  Ignacio Alcantarilla
   
• EUSPA Technical Officer:
  Silvia Porfili

Background

The use of a European Global Navigation Satellite System (EGNSS) – EGNOS and/or Galileo -receiver in combination with other sensors could result in the provision of accurate and reliable positioning which would translate into an overall improvement of the European rail network.

Galileo is Europe’s Global Navigation Satellite System (GNSS) providing an accurate, guaranteed global positioning service under civilian control. It is operational since the Initial Service declaration at the end of 2016. As part of the main services provided by Galileo, the Galileo High-Accuracy Service (HAS) provides (since January 2023) high-accuracy positioning service and synchronisation information.

EGNOS, the European Satellite-Based Augmentation System (SBAS), includes a Safety-of-Life (SoL) service providing the necessary level of integrity, continuity and accessibility to meet the needs of safety-critical applications. In the future, EGNOS V3 will also provide a Dual-Frequency Multi-Constellation (DFMC) service based on augmentation of both GPS and Galileo constellations on the L1/E1 and L5/E5a signals with improved levels of performance. EGNOS V3 will also maintain the legacy augmentation service of GPS L1. The EGNOS SoL service was designed to support civil aviation operations. However, the SoL service is also intended to support applications in a wide range of other domains such as rail, maritime, road... In particular, due to the challenging environments in rail operations (e.g. tunnels, urban canyons, dense vegetation…) which cause masking, multipath, and non-line-of-sight effects, the EGNOS legacy service cannot always provide the required level of performance and, therefore, railway applications would greatly benefit from a multi-constellation approach.

Project objectives

The main tasks of the project were the following:

• Identify what the current safety, legal and regulatory constraints as well as the standardisation framework in the rail sector are at European level.
• Analyse the main technologies used nowadays for Positioning, Navigation and Timing (PNT) in the rail domain and try to predict their evolution, with respect to EGNSS penetration in the rail sector, within the timeframe 2022-2035.
• Develop a new integrity concept for the OnBoard Unit (OBU), to be nested within the ETCS (European Train Control System), for safety critical applications and define the associated requirements that the GNSS receiver and antenna of an OBU shall comply with. Determine the minimum requirements that shall be verified during a test campaign to validate the new integrity concept. 
• Characterise, in terms of performance, an ERSS (EGNSS Rail Safety Service) tailored to the rail domain considering several scenarios and operational conditions. Two service definitions will be considered:
  • the short-term, including the replacement of physical balises with virtual balises, and
  • the long-term, with the objective to improve the overall performance of the safety service to address more demanding rail applications. 
• Estimate how the new proposed service offered by the EGNSS to the rail sector will be beneficial to: 
  • GNSS receiver manufacturers;
  • Rail users: infrastructure managers, rail operators, …;
  • EGNSS safety service provider.
• Assess the different steps to implement the new EGNSS safety service for rail, identifying risks and associated mitigation actions as well as key decisions to be taken. 
• Support EC in the dissemination activities.

Results

Context of the Project

Four applications for the EGNSS safety service for rail were studied in the frame of IMPRESS project: 

• Virtual Balise;
• Enhanced Odometry;
• Track Identification;
• Train Integrity and Length Monitoring.

IMPRESS addresses these applications because they have been identified as the most relevant ones in the frame of the ERTMS (European Rail Traffic Management System) implementation. They are rail safety critical applications and they are dependant of the ERTMS development and roll-out. Moreover, they all clearly justify the penetration and the implementation of GNSS in the rail domain in order to enhance existing applications.

Two scenarios were analysed in detail:

• A short-term scenario with a target time for operational use of the EGNSS Rail Safety Service (ERSS) of pre-2030. In this scenario, the only application considered is the Virtual Balise. Indeed, that application has been developed in order to minimise changes to ERTMS specifications/standards. It could therefore be implemented and used relatively soon, i.e. prior to 2030.
• A long-term scenario with a target time for operational use of the ERSS of 2030 – 2035. In this scenario, it is assumed that the applications for Enhanced Odometry, Track Identification and Train Integrity and Length Monitoring are all applicable. At the same time, the service would keep providing the Virtual Balise application which is assumed remain relevant for some European lines/networks in the long-term. 

Additionally, a very-long-term service was analysed at high-level based on operational use of the ERSS beyond 2035.

Prior to defining integrity concepts for the short-term and for the long-term, application-level requirements in terms of accuracy, availability, availability of integrity, Time To Alert (TTA) and safety integrity level, have been assessed for each of the safety critical application listed above.

As example, in terms of accuracy, virtual balise application requires an ATPE (Along-Track Protection Error) below 15 metres, 99.7% of time and the Train Integrity and length monitoring application a horizontal accuracy between 1 and 10 metres. As for track identification, the horizontal accuracy requirement is lower than 1.9 metre for track discrimination or 2.25 metres depending on the inter-track distance. All applications are SIL4 (Safety Integrity Level), the most stringent one; the TTA (Time to Alarm) is between 6 to 10 seconds for Virtual Balise and Enhanced Odometry while it goes up to 10 to 30 seconds for the two other applications.

On top of that, all the results have been validated by rail stakeholders gathered with the ERGO (Panel of Experts in Rail for EGNSS Operational use) working group. This working group of rail experts has been created to assist the EUSPA in consolidating and validating an EGNSS-based Safety Service concept tailored to the rail community and in line with the ERTMS standards.

Definition of the integrity concept

For the short-term scenario, the integrity concept is based on the existing legacy EGNOS system and service (augmentation of GPS L1 only) and on the provision of performance commitments in the pseudorange domain excluding local environment effects. Additional trackside augmentation data distribution function has been considered as part of the project to enhance the reception capabilities of the ERSS data additionally to the legacy dissemination through Signal in Space (SiS) of the geostationary (GEO) satellite and then provide it to the users (Onboard) over a safe radio channel. The development and implementation of this function is envisaged to be railway responsibility, e.g. infrastructure managers.

For the long-term scenario, the integrity concept is based on the Dual Frequency Multi Constellation (DFMC) EGNOS system and service (augmentation of GPS L1, L5 and Galileo E1, E5a) and a similar service concept to the short-term one: to provide performance commitments in the pseudorange domain excluding local environment effects. Despite this similarity, there are significant differences between legacy and DFMC EGNOS and in particular the way ionospheric delay is handled. It is also expected that, thanks to the improved performances of DFMC, the long-term scenario will offer improved performance levels with respect to the short-term scenario. Finally, for the long-term scenario, a dedicated terrestrial channel between the EGNOS ground segment and the Trackside, will replace or complement the aforementioned use of the GEO SiS.

Where along-track positioning (1D) is required, it is recommended to adopt a tight integration/map-aiding approach due to availability and integrity performance benefits.

Protection level equations are proposed for both scenarios. In the short-term scenario, an along-track protection level is computed based on the used ETCS odometer coasting method. In the long-term scenario, the equation varies depending on whether horizontal or along-track positioning is computed (which is application-dependent) and whether coasting is being performed using the ETCS odometer or carrier phase measurements. 

The recommended approaches to coasting involve using the ETCS odometer in the short-term scenario, and a hybrid method in the long-term scenario that would use the odometer and carrier phase measurements (including the use of Relative RAIM, RRAIM, to provide integrity during coasting). IMU (Inertial Measurement Unit) technology is recommended to improve integrity, including mitigation of feared events caused by the local railway environment and GNSS receiver, and to provide a solution when GNSS cannot be used (e.g. in tunnels). Finally, a set of new EGNSS mission requirements were specified, derived from the integrity concept analysis and existing EGNOS requirements.

Service Definition

Following the definition of integrity concept, the IMPRESS project proposes Service Definitions for the EGNOS for Rail Safety Services based on the defined integrity concepts (short and long term) to be provided by the EGNSS. Additionally, a very-long term scenario (beyond 2035) was analysed at high level and an associated Service Definition was proposed.

For short term scenario, only one service called “single frequency ERSS” was defined. The single frequency ERSS is foreseen to be a by-product of EGNOS V2 and EGNOS V3.1 SoL service. The service dissemination would be done through the EGNOS space segment using the aviation SoL messages (NOF). There are no additional messages. A new standard, similar to the aviation ICAO SARPs, specific to the railway domain is envisaged, that will specify and contain requirements for the performance in pseudorange domain needed for ERSS, for the definition of the additional delivery channel. Such SARPs like standard will be an applicable document to the envisaged service.

The required service performances are derived from the mission requirements that are linked to pseudorange level performances, for which a commitment is needed at EGNOS level. 

As for the safety and regulatory requirements, ERSS being considered as a supporting service based on a pre-existing system, no additional development of safety case is deemed necessary. Nonetheless, a statement of supporting evidence (EN50129/IEC61508) as a Safety Manual as defined in IEC61508-4 has to be developed by the service provider.

The EGNOS service provider will be required to hold a certificate that covers the part of the infrastructure as part of the provision of the ERSS. This certificate will have to be granted by the Member State in which the service is provided first.

For long term scenario, there are two additional services proposed on top of the single frequency service (described above): 

• Dual Frequency Multi-Constellation ERSS, 
• Authentication service. 

The dual frequency multi-constellation ERSS service is a side-product of EGNOS V3.2. The authentication service would be delivered to the trackside equipment through a dedicated dissemination channel using E5b. It is proposed to add the authentication system functionality and the additional dissemination channel to EGNOS system. The authentication service level authenticates EGNOS messages for both single and dual frequencies service levels. In terms of standards, additional aviation like SARPS should be developed to cover:

• the requirements in pseudorange domain needed for DFMC ERSS,
• the definition of the additional delivery channel,
• authentication data definition

No additional development of safety case is required as detailed for the short-term scenario above. 

The very long-term scenario addresses the provision of a service based on EGNSS High Accuracy Service (HAS) SL3, including integrity data. This service would be forecasted in the 2035 + timeframe. In the scope of very long-term scenario, ERSS would provide one new service called here “High Accuracy service”. The services described for long term would still be available.

Decision Criteria

The third step of the project was to evaluate the implementation of the scenarios (short-term and long-term) from the perspective of the three different stakeholders (railway operator, infrastructure manager and device manufacturer) and at industry level. 

The long-term scenario was divided into two scenarios: 

• Long-term scenario A covers the implementation of enhanced odometry and track identification services.
• Long-term scenario B covers the implementation of enhanced odometry, track identification and Onboard Train Integrity (OTI) services. The scenario is independent and does not rely on the previous implementation of either the short-term scenario or long-term scenario A.

All those scenarios (short-term, long-term A and long-term B) were compared to a reference scenario that considered the continued, linear deployment of ERTMS in Europe, without the implementation and benefits of either the short- or long-term scenarios.

The following assumptions were made in the calculation of the Net Present Value (NPV) and return on investment: 

• Discount rate: 8%; 
• Base year: 2021;
• CBA (Cost Benefit Analysis) horizon: 2040.

Concerning the main assumptions taken for the short-term scenario:

• Deployment of Virtual Balise from 2024, with service becoming operational from 2027;
• Virtual balise to be deployed on all ERTMS upgrades of conventional mainline railways and on newly built mainline railways between 2024 and 2040;
• 75% of physical balises to be replaced by virtual balises on those railways;
• By 2040, 50% of all railway lines equipped with virtual balise, that is 72% of all railway lines with ERTMS;
• All locomotives and multiple units to be equipped with the corresponding onboard units (OBU).

As for the long-term scenarios (A & B), the following hypotheses were made:

• Deployment expected to start after the TSI milestone and with availability of EGNOS DFMC in 2028, and services becoming operational from 2031;
• Enhanced Odometry and Track Identification to be deployed on all ERTMS upgrades of conventional mainline railways and newly built mainline railways from 2028
• Enhanced Odometry to be deployed on 38% of all railway lines, or 54% of all railway lines with ERTMS by 2040;
• All locomotives and multiple units to be equipped with the corresponding onboard units (OBU);
• By 2040 75% of all locomotives / multiple units’ system-wide to be equipped.

The availability of ERTMS Level 3 would be required and specific to long-term B scenario.

The assessment of each implementation scenario, from the perspective of the three different stakeholders and at industry level, shows that, for device manufacturers, the costs of foregone sales profits (physical balises or axle counters) (depending on the scenario) is partly offset by the sale and installation of onboard units. Depending on the scenario, these incremental sales can exceed foregone costs/profits (in the short-term scenario and long-term scenario A). In all scenarios, infrastructure manufacturers benefit from CAPEX savings, being able to save on physical balise, train integrity equipment and related implementation costs. The costs related to equipping vehicles can become an important burden for railway operators, preventing these from experiencing Net Positive Benefit under all three scenarios. Infrastructure managers and railway operators are likely to benefit from additional capacity that can be created with ERTMS L3 and onboard train integrity (OTI) under long-term scenario B, by far offsetting the costs incurred. Such a positive outcome only applies if the deployment of the OTI service takes place on railway lines with high demand, meaning that there is a positive business case for railway companies and infrastructure managers to operate additional services and generate income. 

Table 1 summarises the results of the cost-benefit analysis performed for each implementation scenario and for each stakeholder. 

Table 1 - CBA Results

By assessing the financial viability of the three proposed scenarios for EGNSS applications in railways, the CBA developed in this project provided reference key decision criteria for taking the deployment forward. The criteria can be summarised as follows: 

• The NPV of the deployment scenario needs to be positive.
• The cost burden should be fairly split, whereby costs and benefits born by each stakeholder should be proportionate. If the CBA shows a strong imbalance, the redistribution of costs and benefits should be considered.

As shown in Table 1, the financial results of the CBA for the three scenarios are characterised by extremes: two scenarios turn yield a negative NPV, and one yields a positive NPV. Furthermore, the net impact on individual stakeholder groups is distributed unevenly, with railway operators generally bearing the main cost burden, and device manufacturers or infrastructure managers benefiting most. 

• Short-term scenario and long-term scenario A not viable

The short-term scenario and the long-term scenario A, both, have a negative NPV: railway operators have to bear the high costs of having to equip their fleets with OBUs; device manufacturers benefit from a moderate net increase in sales profits, and infrastructure managers enjoy a reduction in capital expenditure. Still, the combined financial and operational benefits across all stakeholders are insufficient to support the aggregate costs. As a result, both scenarios have a negative NPV. 

• Only Long-term scenario B is viable

Long-term scenario B has an overall positive NPV due to the high benefits for infrastructure managers. 

Criteria for assessment: 

• Clear benefits and overall positive NPV;
• No single stakeholder group should support an over-proportional share of the overall costs. 

Under these criteria, only the long-term scenario B qualifies from a financial perspective, under the condition that an alternative cost-benefit distribution can be agreed. This could involve a reduction in network access charges, meaning that infrastructure managers share part of the benefit with railway operators, who, in turn, would be alleviated from the cost of OBU deployments (and hence be more likely to invest in OBU, for the benefit of all). 

Given the results of the CBA, the short-term scenario and long-term scenario A are not recommended, unless a reduction in OBU costs or greater benefits for any of the stakeholders can be achieved (and fairly redistributed amongst stakeholders).

Roadmap for service implementation

Both scenarios, short-term and long-term, should be thought of independently.

• On Figure 1, the short-term scenario assumes the ERSS rollout to occur at the earliest feasible opportunity, i.e. before 2030 (a target date defined by the IMPRESS study). Looking forward, the short-term scenario anticipates reliable results from the demonstrator projects which should culminate in the finalisation of a draft change request to the existing CCS TSI (Control Command and Signalling Technical Specification for Interoperability) by the industry by 2024, e.g. based on the X2R5 WP05 outcome (cf https://projects.shift2rail.org/s2r_ip2_n.aspx?p=X2RAIL-5). From 2024 to 2025, ERA, at the receiving end of the draft proposal, would focus on concluding the necessary steps for adopting a modified TSI in 2026. Subsequently, the following activities would look at the regulatory definition of the service, the provision of support for the industrialisation (including the financing of pilot projects), and the development of a detailed safety case for specific solutions (including the application design and the physical implementation) developed by suppliers and manufacturers. The efforts deployed during this period would allow for an entry into service of the ERSS, including a first rollout of equipment in late 2028/early 2029

Figure 1- Roadmap for the short-term scenario

EGNOS V2 was initially created for aviation, and more specifically for precision approaches like CAT-I LPV200. Its performance requirements were defined in for the position domain, which worked well for aviation applications. However, for railway applications, the integrity in the pseudorange domain (the calculated distance from a satellite to the receiver) is essential, but it was not originally specified in EGNOS V2. This means additional work is required to adapt EGNOS V2 for railway use. This work involves developing more detailed requirements related to pseudorange integrity, analysing potential issues (both expected and worst-case scenarios), and devising solutions to these issues to ensure a suitable level of pseudorange integrity for railways. This will have to be done as part of Task 2. On the other hand, EGNOS V3.1 already includes requirements for pseudorange domain integrity. Therefore, it can commit to performance in the pseudorange domain. This means that, with EGNOS V3.1, a service declaration for railway applications can be made based on pseudorange integrity, even without considering local environmental effects. 

Several steps need to be taken to enable cross-acceptance by the railway domain of the existing EGNOS safety case. In the case of EGNOS V2, specifications need to be developed into requirements on range domain integrity as noted above. Moreover, the establishment of EGNOS working agreements (EWAs) to formalize the operational and technical modalities for using EGNOS within ERTMS will need to be addressed. 

Finally, the EGNOS system would need to be certified to meet the safety requirements of the railway domain, which would involve obtaining safety certification from the European Union Agency for Railways (ERA) and ensuring compliance with the Common Safety Method for Risk Evaluation and Assessment (CSM-RA). The approach to service agreements between the ERS Service provider and the railway domain is expected to be similar to the approach used in aviation. The ERS Service provider would need to establish a working agreement directly with an agency or party, yet to be identified, to formalize the operational and technical modalities for using ERSS for railway applications. This will be part of Task 3. This would involve providing commitments on the ERSS service and defining service arrangements between the ERSS service provider and the counterpart for the use of ERSS within ERTMS, including EGNOS data recording, collaborative decision-making, and involvement of the counterpart in the ERSS decision-making process where decisions could lead to a material impact on the service provided. The Infrastructure Managers (IMs) would be involved in the process through the provision of ERTMS System Compatibility (ESC) and Radio System Compatibility (RSC) parameters classifying their trackside network, which are required to demonstrate technical compatibility. Specific parameters indicating that the infrastructure provides the "GNSS augmentation service" could also be considered. Receiver guidelines based on aviation MOPS have been already drafted as part of the CR1368. These are currently being further matured with feedback from X2R5 consortium members. This applies to legacy (GPS L1) and DFMC modes of EGNOS.

On Figure 2, the long-term scenario takes into account the evolution of EGNOS over the upcoming years, with a specific focus on the pivotal role of EGNOS V3.2, set to declared in 2028. The long-term scenario, commencing with an alignment to the short-term plan, leads to a revision of the CCS TSI in 2026, underpinned by insights from ongoing demonstrator projects. We then propose a further modification to the 2029 CCS TSI, aiming to incorporate new functionalities such as EGNSS-based Enhanced Odometry and Track Identification. Concurrently, we propose to investigate essential modifications to the regulatory definition of the service, facilitate industrialisation, and construct a comprehensive safety case for solutions derived from the 2026 CCS TSI. Considering the future rollout of ERSS post-2030, we believe this timeline may be more realistic if a consensus is reached among stakeholders regarding the necessity of having DFMC available by 2032 for service deployment. This could be essential in addressing persistent performance gaps and challenges in complex operational environments like urban canyons or tunnels, where DFMC benefits would be notably impactful. Subsequently, factoring in the anticipated three-year period between the 2029 CCS TSI rollout and the operation of DFMC-based ERSS, we predict that DFMC-based ERSS would be fully operational by 2032.

Figure 2 - Roadmap for the long-term scenario

The long-term scenario considers the evolvement of EGNOS over the next years, with the launch of the service depending on the important contribution of EGNOS V3.2 available by 2028. 

The roadmap begins with the implementation of demonstrator projects between the present day and 2024. These projects will test the viability of EGNSS, and the results will be used to propose changes to the TSI in 2024. Once these changes are proposed, the European Union Agency for Railways (ERA) will action them, leading to modifications in the CCS TSI in 2026.

In parallel, between 2024 and 2026, stakeholder consensus will be built around the importance of the DFMC, set to become available in 2028, for the successful rollout of the service foreseen in 2032. In particular, it will be argued that DFMC is critical to overcome challenges faced in complex operational environments.

From 2026 to 2028, this period will also see the regulatory definition and setup of the service in anticipation of the EGNOS V3.2 launch in 2028. The next milestone will be the implementation of additional functionalities (like Enhanced Odometry, Train Integrity and Monitoring and Track Identification) in the CCS TSI during 2029. Then, from 2030 to 2032, a testbed will be set up, and the development and validation of equipment will take place.

Industrialisation process will start straight after the publication of CCS TSI in 2029. After a period of post-launch evaluations and adjustments from 2030 to 2032, EGNSS deployment is expected in 2032, following a development time of three years post the 2029 TSI change. Finally, in 2032, the ERSS will be formally launched, utilising the DFMC to enhance system performance and address any remaining performance gaps.

Figure 3 illustrates the critical path of the roadmap:

Figure 3 - Critical Path of the IMPRESS Roadmap

Expected impact

IMPRESS set up the basis of an integrity concept for an EGNSS-based rail safety service for some rail critical applications: Virtual Balise, Enhanced Odometry, Track Identification and On-Board Train Integrity.

An ERSS has been defined based on EGNSS requirements, for a short-term (pre-2030) and for a long-term scenario (btw 2030-2035).

The study analysed the costs and benefits of the solution from several points of views and the results are driven by capacity-generated and the cost of the on-board unit; the Long-Term B scenario is the only financially viable.

Two roadmaps have been charted, envisaging a Virtual Balise roll-out before 2030 in the short-term and the four applications deployed between 2030 and 2035 in the long-term.

There are still several open points:

• Rigorous derivation of integrity budget allocations, in particular for the On-Board Train Integrity, Train Integrity and Track Selectivity
• Performance of the proposed concepts in terms of ‘availability of integrity’ (protection level magnitudes).
• The entity that will provide the ERSS
• The denomination of the envisaged standards in this document.
• The entity that will be responsible for the drafting of such standards, and their endorsement
• The exact coverage of EGNOS V3.2 that will determine the coverage of the ERSS in the long-term scenario
• The exact timeline of the CR 1368 publication and when the next TSI will exactly be.
• The entity counterpart that will be in charge of signing the EGNOS Working Agreements to formalize the operational and technical modalities for using EGNOS within ERTMS
• The cost of OBU under the short-term scenario
• The capacity gain under the long-term scenario.

Disclaimer

The project results represent the views of the consortium. They do not necessarily represent the views of the European Commission and they do not commit the Commission to implementing the results.

Consortium

IMPRESS Consortium is composed of Egis Avia, which leads the activity, Egis Rail, Thales Alenia Space (leader), and GMV UK.

The IMPRESS consortium would like to warmly thank all of the experts from EUSPA and from ERGO (Experts in Rail for EGNSS Operational use) who contributed to the review of the project results.