Cross Sectorial Task Forces Archives - BDV Big Data Value Association

Emerging Topics

nimdabdva — Sat, 23 Sep 2023 07:34:15 +0000

It has the charter to cluster new and promising activities that span several technical domains and can span important advantages to SMEs and industry. For this, the Emerging Topics Task Force will constantly survey the industrial and academic ecosystem for new promising technologies that have the capability to impact the business of European companies. The current focus areas are two: the Metaverse on the one hand and Neuromorphic computing on the other hand, but the TF is also open to exploring other emerging topics of interest.

BDVA’s Task Forces meet regularly, discussing the latest updates from the ecosystem, collaborating on publications and organising events. Check out BDVA’s activities and make sure to stay tuned!

The post Emerging Topics appeared first on BDV Big Data Value Association.

Generative AI / Foundation Models

nimdabdva — Sat, 23 Sep 2023 07:29:53 +0000

The rise in generative AI, underpinned by new foundation models, offers numerous opportunities for businesses and society in Europe. The EC AI innovation package launched in January 2024 includes the GenAI4EU initiative, which aims to foster the development of new use cases and emerging applications. Industries and the public sector will require support to navigate this rapidly evolving landscape and effectively leverage generative AI within their organisations.

The challenges and opportunities presented by generative AI (GenAI) span across various domains, necessitating a comprehensive approach to address them effectively. One critical area is the need for a thoroughly developed infrastructure to train and operate foundational models, which can be facilitated through collaborations with programmes and the AI Factories through the Task Force HPC-BD-AI and our participation in the EuroHPC JU.

Additionally, GenAI workloads are known for their significant energy consumption, highlighting the necessity for optimisation strategies to mitigate their environmental impact.

Access to data, along with security and privacy, is paramount, including evaluating and countering systemic GenAI effects, such as the proliferation of synthetic data and the rise of cybersecurity threats like automated phishing and deep fakes. Another crucial aspect is the rigorous testing of both foundational and generative AI models to ensure their reliability and effectiveness.

Identifying the most accessible and impactful business or industry use cases for GenAI is essential, as is understanding and addressing the barriers and challenges to their adoption. The successful implementation and operation of foundational models and GenAI require specific technical skills, including the incorporation of human-in-the-loop systems.

Ensuring the explainability of foundational and GenAI models is critical for transparency and trust. Compliance with ethical standards and regulations, such as the AI Act, is necessary, achievable through collaboration with Task Force etami. Finally, providing guidance on the future stages of GenAI evolution, addressing forthcoming challenges and leveraging insights for technical and business strategies are vital for sustained progress in the field.

The Task Force Generative AI has set several medium- to long-term objectives to advance its mission. One key goal is to establish a forum that addresses the previously identified challenges and explores their impact on the practical implementation of Generative AI. This forum will foster collaboration, co-creation and the formation of partnerships, particularly in the realms of research, development and deployment of Generative AI and Foundation Model-based solutions, including new proposals for Horizon Europe.

Additionally, this task force aims to support European Commission initiatives like GenAI4EU and synthetic data projects by representing the viewpoints of BDVA members. It will also generate strategic content to positively influence stakeholders and assist members in adopting Generative AI technologies. This support will encompass both implementation and usage, covering knowledge about various foundation models and the need for new, specific industrial models.

This task force will collaborate closely with other BDVA task forces that intersect with Generative AI and Foundation Models. This includes both sector-specific areas, such as healthcare and cross-sectorial areas, such as high-performance computing (HPC), data spaces, etami, data protection and the trustworthiness of industrial AI, ensuring a comprehensive and unified approach to advancing AI technologies. The task force will also seek collaboration with existing partnerships and programmes, notably with Adra.

The BDVA Task Force Generative AI / Foundation Models is currently focused on:

To identify and analyse real-world applications of GenAI across industries, understanding how to assess its impact, scalability and business potential.
To develop strategies to facilitate industrial stakeholders on leveraging GenAI providing a framework for guidance and training to empower businesses, SMEs, and public institutions to adopt GenAI responsibly and efficiently as well as handling the AI literacy requirements in the AI Act.
To facilitate the adoption of GenAI by companies through practical implementation support. Many companies struggle with the integration of GenAI into their workflows, therefore practical tools and guidance for seamless adoption are being discussed in the scope of the task force.

The post Generative AI / Foundation Models appeared first on BDV Big Data Value Association.

Standards and Benchmarking

nimdabdva — Sat, 23 Sep 2023 07:29:28 +0000

The Standards and Benchmarking Task Force focuses on the rapid evolution of international standardisation activities in Big Data, Artificial Intelligence, Data Governance and Robotics and on facilitating international trade, business scaleup, emerging technologies and their sustainable development.

For the use of AI solutions in industry and academia, benchmarks and key performance indicators are central as they allow the data-driven comparison of the results of corresponding algorithms. Addressing these challenges at a European scale is central for the development of competitive benchmarking solutions that can be used to advance AI in Europe and the world.

The BDVA Task Force on Standards is currently focused on:

Standardisation in data and data spaces, leveraging the knowledge of the community in this field, the participation of BDVA in CEN Trusted Data Transactions CWA, its liaison with CEN CENELEC JTC25 Data, dataspaces, cloud and edge, and the partnership in Data Spaces Support Center standards WG

Standardisation in AI, capitalising discussions from other Task Forces and funneling this knowledge to CEN CENELEC JTC21 Artificial Intelligence and ISO/IEC JTC1/SC42 Artificial Intelligence

Standardisation in virtual worlds, combining the knowledge generating in the Task Force Emerging Topics, and the participation of BDVA in the project OpenVerse, to evaluate existing standards on the different technologies related to virtual worlds, and identify needs in their landscape

Standardisation in digital twins, since the topic aligns data-driven innovation, AI, interoperability, and industrial transformation, requesting interoperability between multiple fields, and being standardisation the way to scale-up in the topic. BDVA will leverage the knowledge of its community and the participation of key members in ISO/IEC JTC1/SC41 Internet of things and Digital Twin

The post Standards and Benchmarking appeared first on BDV Big Data Value Association.

Skills

nimdabdva — Sat, 23 Sep 2023 07:29:03 +0000

The Task Force Skills focuses on the growing pervasiveness of data in business and society that has made a central need for data scientists to help organisations extract knowledge from their data. Filling this skills gap in Europe is a challenging task that has many facets, such as:

Educators (from academy to any organisation offering courses) need to know the needs and fills the gaps for different levels of expertise (professional/theoretical, formal/non-formal);
Employers needs to hire/retrain personnel with the proper skills, depending on the kind of organisation (SME/large industry) different approaches are possible;
Employees and students need to get their skills recognised;

The post Skills appeared first on BDV Big Data Value Association.

Policy and Societal

nimdabdva — Sat, 23 Sep 2023 07:28:53 +0000

The Task Force Policy, Societal and Regulation contributes to BDVA by addressing not only technological developments in Big Data, but also identifying a number of topics that are policy-related, legal, or societal and linked to such technological developments. Hence, while data-driven innovations hold great promises for the advancement of society, a growing debate on non-technological aspects of these innovations can be observed.

This debate drives policy-making, legislation and public opinion. The key to responsible research and innovation is a dialogue between researchers, technology developers, policy makers and end users. Thereby, this TF follows closely the policy-making agenda of the European Commission, with a particular focus on the Data Strategy, the AI Strategy and the Path to the Digital Decade Strategy (2020 – 2030).

The post Policy and Societal appeared first on BDV Big Data Value Association.

i-Spaces

nimdabdva — Sat, 23 Sep 2023 07:28:40 +0000

European Data Innovation Spaces, or i-Spaces in short, have been a fundamental part of the Big Data Value Association’s story from the very beginning. Early on, the BDVA community realised the need for a tool to support businesses and researchers in order to extract value out of their data. For this purpose, BDVA created the i-Space label, to distinguish data-driven innovation hubs (DIHs) with the latest knowledge, skills, data assets and software. By federating these qualified DIHs, BDVA enables the creation of a flow of data services to boost innovation in EU member states and lays a path towards European data spaces.

i-Spaces are giving continuity to the Federation created during the EUHubs4Data project. The Federation facilitates cross-border and cross-sector data-driven experimentation through collaboration between hubs, data sharing and joint services delivery. To establish the collaboration, the Federation has created a common catalogue which includes services, databases and courses offered by the members. A strong commitment to ethics has always been at the heart of the project. For this reason, the Federation developed an Ethical Toolkit that enables the assessment of organizations from an ethical perspective.

The idea behind the i-Spaces is explained in details in Vision Paper of the European Federation of Data Innovation Spaces (i-Spaces) available here.

You can learn more about the i-Spaces here and the current list of i-Spaces is available here.

Access the Catalogue

The post i-Spaces appeared first on BDV Big Data Value Association.

HPC – Big Data – AI

nimdabdva — Sat, 23 Sep 2023 07:28:31 +0000

The Task Force focuses on leading several research and innovation activities relevant to both EuroHPC JU and the AI, Data and Robotics Partnership, supporting the alignment between both partnership communities with the goal of developing common business and research opportunities.

The BDVA Task Force on HPC/Big Data/AI is currently focused on:

AI Factories, supporting their access to data and connection to data spaces, the implementation of a networking layer that homogenises AI Factories at user level, the assessment of how the different stages of the Generative AI lifecycle are supported by AI Factories, and fostering the onboarding by industry by identifying their specificities and understanding their needs

AI workflows and data pipelines over HPC, bringing together AI practitioners, data experts and HPC infrastructure providers, to assess how to enhance AI performance and scalability in high-performance computing (HPC) environments

AI Software for HPC (and Quantum), by creating a dedicated expert group to work on advancing AI software frameworks and tools tailored for HPC and quantum

Computing continuum, continuing the collaboration within the Computing Continuum Initiative and addressing the main challenges identified in the corresponding BDVA Strategic Theme “Making data and AI innovations fit for emerging infrastructures and platforms”

The post HPC – Big Data – AI appeared first on BDV Big Data Value Association.

etami

nimdabdva — Sat, 23 Sep 2023 07:28:14 +0000

etami focusses on making AI a reliable and trustworthy technology. It does so by focussing on the social, legal and business risks associated with AI systems, rethinking AI lifecycle models to optimise quality in AI, and promoting auditing throughout the system’s life cycle. etami’s solutions are piloted across different domains, and coupled with proper software tooling.

There are many challenges on this path. For instance, questions around fairness and non-discrimination, at the heart of trustworthy and ethical AI, are yet to be fully understood in the current literature, let alone deemed solved from a corporate process’s standpoint. Human oversight is technically challenging, while data documentation tools are yet to be established.

etami started in 2019 as an independent initiative. BDVA, with its large and diverse set of members, helps etami to develop what is crucially missing in the AI landscape: the bridge between solid research and its translation into a reliable and trustworthy engineering practice, while keeping universal ethical values at the centre.

The BDVA Task Force etami is currently focused on:

Explore current trends and key topics beyond the AI Act. It is key to understand the next calls, ideas and trends for Trustworthy AI and Risk management that can have an impact on AI, industry and society
To identify current practices, services that can be used by the industry to support the industry on adoption or to facilitate the analysis and understanding of technology adoption from any of the trustworthy AI dimensions.
To understand the current market status, gaps and opportunities to embrace any of the new technologies or to advance on any of the strategic ones that can bring benefits to industry and society.

Find more information on etami’s website.

The post etami appeared first on BDV Big Data Value Association.

Data Spaces

nimdabdva — Sat, 23 Sep 2023 07:27:53 +0000

The Task Force concentrates on securing data sharing in order to realise the practical potential of sectorial European data spaces.

The efforts of this Task Force are directed towards a cross-cutting work that will enable the creation of Data Spaces, which are a pillar of the Digital Europe Strategy. This Task Force therefore focuses on:

Engaging with all other relevant Task Forces and working groups;
Engaging the broad BDVA data sharing stakeholder community (producers, intermediaries and consumers along data value chains) to identify opportunities, challenges and possible solutions for facilitating cross-sectoral data sharing and exchange practices;
Supporting the EC with mobilising initiatives seeking to realise an adequate European data sharing space for all kinds of stakeholders;
Liasing with key initiatives such as Gaia-X, IDSA, InnoPay, MyData, Datapace.io and other success stories in data sharing spaces or marketplaces;
Engaging this broader community by producing white papers with a state-of-the-art overview (initiatives, projects and solutions) and a vision for how a safe, fair and ethical data-sharing space can be achieved at EU-level.

The BDVA Task Force on Data Spaces is currently focused on:

Data value creation, advancing BDVA’s core mission of fostering innovation, competitiveness, and economic growth through data-driven and AI technologies.
AI-ready Data Product, advancing the data product paradigm to better serve AI practitioners and applications, including methods for optimizing data preparation for AI
Interoperability between data spaces, as a key driver to breaking silos and unleashing the full value of data spaces
AI and Data Spaces, leveraging the vast amount of high-quality data from data spaces to produce and train large AI models, strengthening the connection between AI and data spaces while positioning AI as their key enabler

The post Data Spaces appeared first on BDV Big Data Value Association.

Data protection

nimdabdva — Sat, 23 Sep 2023 07:27:42 +0000

The Data Protection Task Force focuses on the primordial influence of data protection on many aspects involved in data processing, including utility, performance and quality.

This process starts with the growing awareness of the salience of data protection and results in increasing interest in privacy-preserving technologies and other technical and non-technical measures to ensure data protection. Also, the regulatory requirements on data protection have become strict and complex in recent years and this trend will likely continue in the future. There is a growing awareness of the importance of data protection, resulting in increasing interest in privacy-preserving technologies and other technical and non-technical measures to ensure the protection of data. Data protection has become a cross-cutting concern that has important influence on many other aspects of data processing, including utility, performance and quality. This process is further intensified by the upsurge of AI, requiring the processing of ever larger amounts of data and offering new attack possibilities on data privacy.

The post Data protection appeared first on BDV Big Data Value Association.