European Commission logo
English English
CORDIS - EU research results

Programme Category


Article available in the following languages:


Verifiable robustness, energy efficiency and transparency for Trustworthy AI: Scientific excellence boosting industrial competitiveness (AI, Data and Robotics Partnership) (RIA)


Develop trustworthy AI technology, key for acceptance, to take full advantage of the huge benefits such technology can offer, and demonstrate the benefits in particular applications. This will require improvement in transparency: explainability, accountability and responsibility, safety, expected levels of technical performance (accuracy, robustness, level of ‘intelligence’ and autonomy, etc.) which are guaranteed/verifiable and with corresponding confidence levels.

Build the next level of “intelligence” and autonomy, essential to scale-up deployment, in solving wider set and more complex problems, adapting to new situations and context knowledge, addressing real-time performance requirements and data and energy efficiency, also for greener AI and robotics solutions. This will investigate approaches such as integration of both learning and reasoning, causality, contextualization and knowledge discovery, hybrid semi-parametric models (combining laws of physics with observations, aka physics-informed machine learning), human-in-the loop approaches, etc.

Contribute to making AI and robotics solutions meet the requirements of Trustworthy AI, based on the respect of the ethical principles, the fundamental rights, including privacy. Ethics principles needs to be adopted from early stages of AI development and design.

In this topic, solid scientific developments will be complemented, as relevant, by tools and processes for design, testing and validation, certification (where appropriate), software engineering methodologies, as well as approaches to modularity and interoperability, aimed at real-world applications. Where appropriate proposals are encouraged to propose standardisation methods to foster AI industry, helping to create, and guarantee trustworthy and ethical AI, and in support of the Commission regulatory framework.

Scientific proposals are expected to focus on advancing the state of the art in one of the major research areas below:

  1. Novel or promising learning (such as unsupervised, self-supervised, representational learning capable of contextualization, transfer learning, life-long and continual learning, etc.) as well as symbolic and hybrid approaches. The objective is to advance “intelligence” and autonomy of AI-based systems, essential to scale-up deployment, in solving a wider set of more complex problems, adapting to new situations (making them “smarter”, more accurate, robust, dependable, versatile, reliable, secured, safer, etc.), and addressing real-time performance requirements, where relevant, for both robotics and non-embodied AI systems. This will include, among others, integration of both learning and reasoning, combining data-driven and knowledge-based models, causality, contextualization and knowledge discovery. Approaches can build on simulation and digital twins, or include data augmentation, knowledge modelling, federation of AI systems – including the use of distributed data – federated learning, and new AI methods ensuring scalability and re-usability. This topic also supports innovative or promising approaches addressing functional and performance guarantees.
  2. Advanced transparency in AI, including advances in explainability, in transparency (with guaranteed/verifiable levels of performance, confidence levels, etc.), investigating novel or improved approaches increasing users’ understanding of AI system behaviour, and therefore increasing trust in such systems.
  3. Greener AI, increasing data and energy efficiency. This covers research towards lighter, less data-intensive and energy-consuming models, optimized learning processes to require less input (data efficient AI), or optimized models, data augmentation, synthetic data, transfer learning, one-shot learning, continuous / lifelong learning, and optimized architectures for energy-efficient hardware, framework that optimises calculations for energy reduction in big data analytics. This also build on latest results in self-configuring, low-power or energy harvesting capable sensor devices, and low power data transmission and energy reduction in big data analytics (e.g. a framework that optimises calculations, leading to decreasing use of energy, etc.).
  4. Advances in edge AI networks, bringing intelligence near sensors, in embedded systems with limited computational, storage and communication resources, as well as the integration of advanced and adaptive sensors and perception (including multi-modal sensing and active perception, distributed sensing, etc.), but also optimising edge vs cloud AI to maximise the capabilities of the overall system (both globally and for individual users). This builds on latest hardware development (for which synergies with the European Partnership for Key Digital Technologies (KDT) is encouraged), but does not cover such hardware developments.
  5. Complex systems & socially aware AI: able to anticipate and cope with the consequences of complex network effects in large scale mixed communities of humans and AI systems interacting over various temporal and spatial scales. This includes the ability to balance requirements related to individual users and the common good and societal concerns, including sustainability, non-discrimination, equity, diversity etc.

Proposals should clearly identify its focused research area among the 5 listed above.

Proposals should include, as appropriate, the development of tools and processes for design, testing and validation, deployment and uptake, auditing, certification (where relevant), software engineering methodologies, as well as approaches to modularity and interoperability.

To complement the impressive progress in developing individual AI algorithms and components, proposals could also address the development of scientific foundations for designing, modelling, analysing, operating, monitoring, integrating, maintaining and extending AI systems.

In all these topics, involvement of multidisciplinary teams and transdisciplinary research, including SSH as appropriate, will be essential. The consortia should involve world-class research labs and top scientists, joining forces to address these major scientific challenges, and they are strongly encouraged to team up with European companies (large and small) representing major industrial sectors for Europe, genuinely interested in S&T progress in these fields, and which consider adoption of AI “made in Europe” key for their competitiveness[[]].

While the proposals should address scientific foundations, relevance to real-world applications should be demonstrated, in particular through use-cases used to demonstrate scientific progress.

All proposals are expected to embed mechanisms to assess and demonstrate progress (with qualitative and quantitative KPIs, demonstrators, benchmarking and progress monitoring), and share communicable results with the European R&D community, through the AI-on-demand platform[[Initiated under the AI4EU project and further developed in projects resulting from H2020-ICT-49-2020 call]], a public community resource, to maximise re-use of results and efficiency of funding.

Activities are expected to achieve TRL 4-5 by the end of the projects.

Proposals should foresee activities to collaborate with projects stemming from topics relevant to AI, Data and Robotics, primarily in destinations 3, 4 and 6, but also in other destinations and clusters (in particular Cluster 3 on cybersecurity where relevant), and share or exploit results where appropriate.

This topic implements the co-programmed European Partnership on AI, Data and Robotics.

All proposals are expected to allocate tasks to cohesion activities with the PPP on AI, Data and Robotics and funded actions related to this partnership, including the CSA HORIZON-CL4-2021-HUMAN-01-02.