Objectives

With the increasing dependence on data for machine and deep learning –based systems and decision-making in various fields such as healthcare, industry, and transportation, there is a pressing need to develop novel models, abstractions, and methods that can support the creation and storage of diverse data sets, considering the modern needs for data use (e.g., AI/ML, big data etc.), but also the evolving constraints related to privacy, data sovereignty and security across the full data cycle. PLIADES will address the growing need for sustainable and efficient methods for creating large volumes of data from diverse sources, in diverse data spaces and domains. We will explore innovative, AI-based approaches to data creation, with a focus on sustainability and efficiency, in such ways as to facilitate the full data lifecycle. We will focus on AI-enabled metadata generation and association, while employing smart data compression, filtering and multi-dimensional human factors compensation to different types of data, from robot, autonomous vehicle and machines sensors-based, through to data derived from human –robot interaction, as well as industrial production and healthcare processes. This will involve the use of cutting-edge AI/ML based technologies and the use and extension of European Initiatives, to improve data quality, security, and privacy while reducing the carbon footprint associated with the management and maintenance of large amounts of data.

KPIs:

• KPI-1.1: Introduction of an Extension to IDSA Reference Architecture Model regarding data creation;
• KPI-1.2: 1 white paper regarding best practices for data creation in various domains (mobility, green deal, energy, industrial and health car;
• KPI-1.3: 1 list of specific metadata per domain to be used for AI-boosted brokering between data space connectors;
• KPI-1.4: A knowledge injection tool to enable domain experts to participate in data/metadata creation;
• KPI-1.5: Delivery of Data Quality tool and services over data spaces;
• KPI-1.6: Greener data creation, filtering and compression methods for vehicles and robots reduce energy consumption by 15%

PLIADES aims at addressing the protection of data by advancing various of its aspects such as privacy, security, trustworthiness, and sovereignty. Since the datasets may contain personal information, it is required that the entire architecture of the project’s outputs respect the GDPR provisions by design and by default. The data security will be strengthened by the staggered application of international data security standards such as ISO/IEC 27001:2022 and ISO/IEC 27701:2019 based on the level of security required for each type of data, including e.g., raw sensor data, which are extremely valuable for certain AI/ML-based approaches. A concept for automated cybersecurity checks during operation of the data space ensures that the cybersecurity level is always high.

PLIADES will also include in its framework concepts for data ownership and usage policies alongside components that enable connectivity, interoperability and security for data space creation based on principles of European Initiatives like IDSA, GAIA-X, BDVA and FIWARE. PLIADES evaluates the use of self-sovereigned/decentralized identities for a dynamic management of data control and usage policies. Ultimately, the goal is to build a digital ecosystem that empowers individuals and organizations to make informed decisions about their data, offer the possibility to revert decisions at a later point in time and ensure that their privacy and security are always protected.

KPIs:

• KPI-2.1: Introduce a framework that adheres to current legislation and soa security practices;
• KPI-2.2: Adaptive Secure data space architecture;
• KPI-2.3: Implement a privacy-enhancing data processing framework with privacy-preserving technologies.;
• KPI-2.4: Training program for widespread adoption and usage of the platform.

In order to assure data sovereignty and secure exchange of data, IDS-RAM will be utilized to forward data from the data provider and data consumer using unique Decentralized Identifiers (DIDs) for each interaction between any identity and any service on the network utilising proper identity management systems and distributed agreement schemes (e.g., distributed ledger technologies). On the other hand, an advanced cross-domain centralised service will be developed for enhancing the communication of domain metadata brokers by integrating AI-based techniques such as a Recommendation Engine, Ranking System, and AI-enabled Declarative Querying. The resulted AI- boosted brokers will use extended metadata collected based on the project’s best practices and domain experts’ knowledge to provide a matching of a data consumer with possible data providers across different domains.

KPIs:

• KPI-3.1: Development and open-source availability of an AI-based broker among dataspace connectors;
• KPI-3.2: Introduction of the project’s AI-boosted Broker connector as an alternative to current IDSA Broker in a white paper;
• KPI-3.3: at least 1 scientific publication regarding the AI-boosted Broker.

PLIADES will work towards establishing an integrated, interconnected network of data spaces that can facilitate collaboration and frictionless data sharing between data owners and users. The requirement to integrate the entire data lifecycle into current data reference architectures is one of the major obstacles to creating such an ecosystem. This indicates that the entire data value chain, including data collection, processing, management, sharing, storing and usage, should be considered while developing data spaces. PLIADES will create industry-wide standards, tools, and reference architectures to help accomplish this goal. The creation of linked data spaces in the context of EVs can facilitate the ability to gather and share information about car performance, energy use, and charging infrastructure. This may allow businesses to manage EV fleets more effectively and encourage the creation of new EV-related services. Similar to this, when it comes to industry and service robots, integrating data spaces can allow data owners to share information such as health indicators, rehabilitation activities, and HRI data streams, enabling more efficient and effective HRI processes and fostering the creation of new service robot applications.

KPIs:

• KPI-4.1: Delivery of a new dataspace connector (that will extend existing ones) with enhanced data sharing
• functionalities;
• KPI-4.2: Demonstrate the usage of the dataspace connector in at least 3 diverse domains;
• KPI-4.3: Include project’s new connector to IDSA Connectors report;
• KPI-4.4: Connector to be available as Open-source.

PLIADES will explore novel AI-based methods for processing of large, diverse streams of data that will enable advanced analytics and data-based decision support to users. Furthermore, using novel decentralized data connectors, PLIADES will enable the homogenized access towards diverse data spaces in a coherent way, ensuring trustworthy access to relevant data, sovereignty, resilience and security of the data spaces themselves, while preventing misuse. PLIADES will develop advanced algorithms that can quickly process and analyze data, providing actionable real-time insights and analytics that can be used to drive informed business decisions. Special focus will be paid also to the explainability of the AI models trained on the shared data in various domains, such as mobility, healthcare, and industrial domain.

KPIs:

• KPI-5.1: Application and Demonstration of federated learning approaches on cross-domain data spaces;
• KPI-5.2: Publish at least 2 scientific publications regarding federated learning application in data spaces;
• KPI-5.3: Delivery of a system with XAI and re-training mechanisms plus 1 relevant scientific publication.

The project’s toolboxes and standards will be deployed in six (6) diverse use cases oriented around different data-spaces and their interconnections, such as the mobility data space where data from smart vehicles will be used to improve their ADAS/AD functions and energy management, the healthcare data-space where patient and robot operator data will be used to improve service robots’ HRI effectiveness, or the industrial data space where HRI data will be used to improve the collaboration between human workers and robots. The use cases aim to provide a challenging validation suite involving heterogeneous data creation and processing that will be used to create novel interoperability paradigms across data spaces addressing full data life cycles with continuous testing and extensive pilot trials.

KPIs:

• KPI-6.1: Validation and demonstration of project’s solution in X real-world large-scale trials;
• KPI-6.2: Availability of at least 3 use cases in IDSA Radar as examples of established data spaces.

To achieve this objective, a cross-domain architecture will be designed that enables interoperability between different data spaces. With active Data Governance that complies with the Data Governance Act, this architecture will guarantee cross-domain data exchange and analysis. This architecture’s central services will make it easier to implement specific data management regulations that uphold data neutrality and altruism and to enforce them to the domain data governance authorities of each data space. The suggested remedy will deal with the problems that organizations are currently having with data sharing because of business, technical, and legal limitations. The solution will support the creation of new services, goods, and applications that have positive social and economic effects by enabling data sharing and reuse. By ensuring data neutrality and data altruism, which are becoming more and more crucial in the age of data-driven innovation, the solution will support the ethical use of data.

KPIs:

• KPI-7.1: Definition of >=3 centralized data space services compliant with the Data Governance Act,
• KPI-7.2: Achieve high user satisfaction rates for the intermediation services

PLIADES aims to actively contribute to an integrated and coordinated set of models, methods, and technologies in order to meet this goal. This entails coordinating aims and objectives with other efforts while also exchanging expertise, best practices, and resources. The ability to prevent duplication of effort, cut expenses, and maintain a uniform approach to data management and sharing across diverse projects is one of the major advantages of forging synergies between various initiatives. By cooperating, various efforts can benefit from one another’s knowledge and strengths, maximizing the use of available resources. A coordinated and comprehensive approach to the Common European Data Space can also help to guarantee that the data is shared in a manner that is secure, safe, and respects individual privacy rights. This is especially crucial given the growing significance of data-driven innovation and the requirement to guarantee that data is utilized properly and ethically.

KPIs:

• KPI-8.1: Synergies with at least 3 European initiatives,
• KPI-8.2: Organize workshops with other European initiatives.