Examples of domain-specific use cases | |
|---|
| The common European mobility data space (EMDS) aims to facilitate data access, pooling and sharing for more efficient, safe, sustainable and resilient transport. One of the possible use cases to be developed using the data obtained through the EMDS is Mobility as a Service (MaaS). It is a service that provides users with a comprehensive tool to plan, book and pay for different types of transport based on several criteria and by using a single interface. It helps the user to choose the best way to move from one place to another and keep all the information about the journey in one place. Besides the requirement to comply with the horizontally applicable regulations (such as GDPR), it might be necessary to consider sector-specific regulations in order to comply with certain technical or organisational requirements. One of the most important legislations for the mobility sector is the Intelligent Transport Systems (ITS) directive that aims to make high-quality and timely data available for services, such as multimodal journey planners, navigation platforms, and emergency services. More information about the applicable legal framework can be found here: 2024-03-19-deliverable-d3.1-analysis-report-v3.pdf (mobilitydataspace-csa.eu) |
| European Cancer Imaging Initiative (EUCAIM) provides a robust, trustworthy platform for researchers, clinicians, and innovators to access diverse cancer images, enabling the benchmarking, testing, and piloting of AI-driven technologies. It aims to capitalize on the recent advances and successes of Artificial Intelligence systems in helping medical professionals detect and diagnose cancers. Potential use case developed with data from the data space: Due to the necessity to safeguard the rights and freedoms of respective data subjects, as well as the highly sensitive nature of the health sector, cancer images used to train and/or test algorithms or AI systems will have to comply with the GDPR provisions, especially with regards to the processing of special categories of data. Apart from that, AI developers need to take into account the legal requirements applicable to AI in the healthcare sector, especially the following legislations: The European Medical Devices Regulation (EU MDR) and In Vitro Diagnostic Medical Devices Regulation (EU IVDR) laying down the safety rules for AI in healthcare domain, The Artificial Intelligence Act requirements for high-risk AI systems - applicable to a safety component of a medical device.
More information about the applicable legal framework can be found here: Meszaros et. al., 2023
|
| The research and innovation data space, EOSC (European Open Science Cloud), has a specific objective of deploying core services and expanding the EOSC federation of existing research data infrastructures in Europe. Potential use case developed with data from the data space: One of the use cases enabled by the EOSC is The Blue Cloud Initiative, which was developed as the ecosystem for marine research according to FAIR principles. It benefits from EOSC funding and policies, as well as from using EOSC as a multiplier to reach secondary users. It directly supports the Green Deal Data Space and the Horizon Europe mission ‘Restore our Oceans and Waters’. For developing use cases within the framework provided by EOSC, it is important to look at the research-specific requirements, especially regarding the researchers’ access to data, as mentioned above.
|
| European Language Data Space builds a trustworthy and effective data market for exchanging language resources in the public and private sectors. Potential use case developed with data from the data space: One of the most important use cases is developing LLM technologies using language data from datasets available through the data space infrastructure. LLMs are largely used in generative AI systems, which are a typical example of a general-purpose AI model regulated under the AI Act. Under specific circumstances (listed in Annex III of the AI Act), these models can fall under the high-risk category. Therefore, requirements for general-purpose AI systems and high-risk AI will apply cumulatively.
Due to their adverse impact on fundamental rights and people’s safety, providers of high-risk AI systems have a number of requirements that they need to comply with. One of the key requirements relates to ensuring access by certain actors (such as providers, European Digital Innovation Hubs, TEFs and researchers) to high-quality datasets for training, validation and testing of AI systems within the fields of activities of those actors related to the scope of the AI Act. As stated in recital 68 of the Regulation, “European common data spaces will be instrumental to provide trustful, accountable and non-discriminatory access to high-quality data for the training, validation and testing of AI systems”. In light of this, data spaces can be understood as enablers for compliance requirements concerning quality criteria referred to in Art. 10, paragraphs 2 to 5. These requirements refer to specific technical and governance building blocks offered by DSSC. For example, tracking of data sharing and data provenance in a data space is a key instrument for compliance with data quality requirements set in Art. 10 AIA, such as: Relevant design choices (Article 10(2)(a)); Data collection processes and the origin of data, and in the case of personal data, the original purpose of the data collection (Article 10(2)(b)); An assessment of the availability, quantity and suitability of the data sets that are needed (Article 10(2)(e)); Identification and mitigation of biases (Article 10(2)(f),(g)).
|
| The common European data space for cultural heritage is designed to accelerate the digital transformation of the cultural heritage sector and maximise the positive benefits of accessing and using cultural heritage data. Potential use case developed with data from the data space: One of the possible use cases with the data obtained through Europeana PRO could concern developing an extended reality (XR) game using certain collections or other objects protected as cultural heritage. For this use case, it will be important to assess whether the items or collections (also understood as datasets in this case) used by the developers enjoy copyright protection. Many works shared by cultural heritage institutions are already in the public domain. This means they were never protected by copyright, their copyright protection has expired, or the copyright was expressly waived. In addition, for the possibility to use these items, it is also relevant to check whether the user (in this case a game developer) will be able to use one of the exceptions provided for in the Copyright in the Digital Single Market Directive.
|