About the Project
SYNERGIES confronts pivotal challenges within the CCAM community, such as the absence of interoperable scenario databases, time-consuming and expensive development cycles, and regulatory ambiguities. It achieves this by implementing the Safety Assurance Framework developed in HEADSTART and SUNRISE. SYNERGIES furnishes stakeholders with interoperable, federated scenario databases, incorporating data from Safety Pool Scenario Database™, ADScene, StreetWise, VV Methods, L3Pilot, Hi-Drive, and more. This facilitates standardized processes, streamlines development cycles, and ensures regulatory compliance. To accomplish this, SYNERGIES will culminate in a European platform designed to enhance the development, training, virtual testing, and validation of CCAM systems.
The SYNERGIES Platform comprises a Scenario Dataspace, aligned with Europe’s approach to data sharing and competitiveness, and a Marketplace, ensuring continual updates and Dataspace scalability. Furthermore, SYNERGIES encourages the inclusion of new initiatives into the scenario dataspace by offering the requisite tools and guidance, from data processing and scenario identification to scenario database governance. This presents a unique opportunity to amplify investments in research and development, consolidating Europe’s leadership in CCAM development, all while prioritizing safety and data protection.
Only through a collaborative effort we are able to address this challenge. SYNERGIES brings together more than 30 partners, leveraging from previous research and community to multiply knowledge, skills, and resources. This synergetic approach is completely aligned with the European industrial culture and builds upon the concepts laid down by the EU ADS Act 2022 by providing detailed implementation methods and tools for its realisation for safe deployment of CCAM systems in Europe.
Challenges
Safety of CCAM needs to be ensured, and the main challenge for safety validation is that depending on the ODD many different driving situations and complex scenarios need to be tested and validated. Additionally, the European automotive sector is guided by strict testing and validation rules, which mandatorily impose a thorough evaluation of the possible situations a CCAM system will face (including multi-actor complex scenarios, hazards, unusual situations, and challenging conditions). However, the complexity and uncertainty of the driving environment, and the complexity of the driving task itself, imply that the number of possible scenarios that an Automated Driving System (ADS) may encounter is virtually infinite. Other studies suggest that to test a CCAM system to assess its safety and prove that it is 20% better than human driven vehicle, it needs to be driven for over 11 billion miles. Additionally, Hazard Based Testing (HBT) and socio-technical systems advocate that the number of miles driven alone is not sufficient to judge confidence in CCAM systems. Instead, the crucial aspect is the range and variety of scenarios encountered during testing. The must be on understanding and identifying ‘how a system can fail or misbehave’ and subsequently ensuring it does so in a safe and trustworthy manner. The nature of scenarios is fundamental to an assessment of safety.
Two key challenges exist in scenario identification:
- First, identifying the “right” and “representative” scenarios for the relevant ODD.
- Second, understanding if “enough” scenarios have been identified to ensure safety of the system.
The latter may also need to consider the probability of occurrence of the scenarios. Wider challenges with scenario-based validation include defining the safety risk associated with the scenario and the vehicle response to the scenario. Key aspect of this is the definition of “good behaviour” or pass/fail criteria for the system for a given scenario. Only through a collaborative effort we will be able to address this challenge.
Objectives
The main objective of SYNERGIES is to fully enable development, training, virtual testing and scalable scenario-based validation of CCAM systems by empowering the Safety Assurance Framework developed in SUNRISE and by
- providing a unified access to a European scenario dataspace
- extending the range of relevant scenarios providing a new set of scenarios covering rural and urban environments enabling complex situation description
- developing (semi)automatic AI data processing tools for scenario extraction and
- establishing a marketplace to facilitate accessibility to the toolset for future scenario generation.
To achieve this, several objectives have been identified which will be accomplished with the results from each of the Work Packages of SYNERGIES.
Objective 1: Deliver widely accessible scenarios
Objective 1: Deliver widely accessible scenarios to all stakeholders of the value chain through a Scenario Dataspace by federating existing and newly developed scenario databases, scaling up the SUNRISE federated framework. A key element of such dataspace is the federated approach which will enable the provision of scenarios from day 1, by leveraging existing initiatives, and enable the scalability by fostering the inclusion of new initiatives into the platform.
Objective 2: Maximise the usability and coverage of the scenarios
Objective 2: Maximise the usability and coverage of the scenarios provided to a variety of stakeholders to ensure the implementation of scenario-based validation. To achieve this, the requirements defined at the start of the project will consider a user centric approach so that the SYNERGIES platform is specifically designed to satisfy the user needs of the most relevant stakeholders. Additionally, SYNERGIES will enable the provision of scenario metadata that allow to better understand the source, representativeness and limitations of the scenarios provided, to maximise its potential use. AI tools will be developed and integrated in the toolchain to automatically obtain scenarios also ensuring the widest coverage as well as their trustworthiness.
Objective 3: Enable the use of heterogeneous and inclusive data sources for the generation of scenarios.
Objective 3: Enable the use of heterogeneous and inclusive data sources for the generation of scenarios. This will be achieved by assessing and exploiting diverse data sources:
- accident data
- vehicle data
- infrastructure data
- drone data, and
- synthetic data (simulation tools, generative AI),
from different European regions, and facilitating interoperable data processing, scenario identification and extraction, and governance of the Scenario Dataspace and underlying scenario databases. SYNERGIES will identify the most relevant data to be collected to obtain the scenarios required for ODD extension. The SYNERGIES Platform includes a Marketplace to provide all the tools needed to obtain scenarios from raw data (and intermediate steps) ensuring interoperability while maximising the use of multiple sources.
Objective 4: Achieve acceptance and upscaling of the proposed solution
Objective 4: Achieve acceptance and upscaling of the proposed solution through an approach that is transparent, holistic, and interoperable with the main previous EU projects (HEADSTART, SUNRISE, L3Pilot and Hi-Drive) and existing scenario database initiatives. The SYNERGIES platform will be built based on the outcomes of the SUNRISE project which counts with an international Expert Network of more than 240 individuals. Additionally, the SYNERGIES Platform will start with access to more than 2M scenarios, 10,000 accidents and 220,000 hours of data recorded on European roads. Then, to maximise the scaling potential of the SYNERGIES Platform, the project will develop and deliver the necessary tools (through its Marketplace) to onboard new initiatives and grow the platform. Hence, the proposed approach will be supported by a multi-disciplinary community of road safety, CCAM developers (OEMs and Tiers included), data processing, testing and simulation experts; simulation tool suppliers; and stakeholders such as policymakers, regulatory bodies, insurers, and consumer organizations.
Objective 5: Enable a solid AI foundation for CCAM
Objective 5: Enable a solid AI foundation for CCAM that includes AI methods such as machine learning and generative AI for both data and scenario generation and enrichment. The AI-supported data generation will complement existing real-world data and those generated through simulation to provide improved data diversity, richness, and completeness to support scenario generation. Similarly, the AI-supported scenario generation will complement existing methods for scenario identification, extraction, and generation to satisfy the scenario database requirement to support scenario-based safety assurance. Additionally, SYNERGIES will deliver specific datasets for AI training and development.