Q&A on HORIZON-CL4-2023 Calls

The AI data Robotics Association

These are the Questions related to the DEP Calls asked at the Info Day and Brokerage event co-organised by Adra, the European Commission and Ideal-IST. As a first step please make sure to read the full call and topic descriptions. All the details can be found there:

  • HORIZON-CL4-2023-DATA-01-02 - Integration of data life cycle, architectures and standards for complex data cycles and/or human factors, language (AI, data and robotics partnership) (RIA)
  • HORIZON-CL4-2023-DATA-01-04 - Cognitive Computing Continuum: Intelligence and automation for more efficient data processing (AI, data and robotics partnership) (RIA)
  • HORIZON-CL4-2023-DIGITAL-EMERGING-01-01 - Novel paradigms and approaches, towards AI-driven autonomous robots (AI, data and robotics partnership) (RIA)
  • HORIZON-CL4-2023-DIGITAL-EMERGING-01-02 - Industrial leadership in AI, Data and Robotics – advanced human robot interaction (AI Data and Robotics Partnership) (IA)
  • HORIZON-CL4-2023-HUMAN-01-01 - Efficient trustworthy AI - making the best of data (AI, Data and Robotics Partnership) (RIA)
  • HORIZON-CL4-2023-HUMAN-01-02 - Large Scale pilots on trustworthy AI data and robotics addressing key societal challenges (AI Data and Robotics Partnership) (IA)
  • HORIZON-CL4-2023-HUMAN-01-04 - Open innovation: Addressing Grand challenges in AI (AI Data and Robotics Partnership) (CSA)
  • HORIZON-CL4-2023-HUMAN-01-03 - Natural Language Understanding and Interaction in Advanced Language Technologies (AI Data and Robotics Partnership) (RIA)

Partnership Questions

Q: Where can I find partner seekers?
In the Funding & Tenders portal, under each topic description, you can find organisations that are seeking partners for that topic, for example, for the DATA-01-02 topic, scroll down to "Partner search announcements" on the topic screen https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/topic-details/horizon-cl4-2023-data-01-02

Q: In HORIZON-CL4-2023-HUMAN-01-04, is it mandatory to have a Japanese or a Korean partner?
A: It is not mandatory.

Q: If a Korean partner joins the proposal does it have to show its own funding already in the proposal or can look for its own funding after the proposal is accepted?
They should provide a commitment that if the proposal is selected the funding will be available. We don’t want to have international organisations participate and later on leave the consortium.

Q: Adding to the question about the collaboration with Japan and SK, what would be beneficial collaborations with peer Associations for the same topics on these countries (Japan and SK)?
A: Where the participation in the consortium would be via the associations and the outreach to their company members and not companies specifically? This would depend on the proposed work. What will be beneficial? if it is to co-develop technologies then you need the specific companies.

Q: Just to be sure, for HORIZON-CL4-2023-HUMAN-01-01 , not including partners from India and/or Canada could involve that reviewers subtract points to a proposal? Or can it be considered as a factor to untie between proposals with the same evaluation?
No, proposals will not be penalised if they do not have partners in the consortium from India/Canada.

Q: For HORIZON-CL4-2023-HUMAN-01-01, could you please confirm that international partners are not eligible for funding?
Yes, international partners will not be funded via Horizon Europe but will have to secure funding from their country.

Questions on HORIZON-CL4-2023-DATA-01-04

Q: Could you clarify the constriction about not using existing technologies. It is hardly possible to innovate a functionality without putting on the current level of technology with cameras, hololense, microphone and some other sensors.
A: The topic is about developing advanced cloud/edge technologies and demonstrating the generic applicability through a number of use cases/applications. If the proposal is about developing a solution/application (i.e. smart city, health care, etc.) using current cloud/edge technologies is not in the scope of the topic.

Q: Is a more direct application intended for industrial installations, factories? Or is it intended to create a system that can more easily support autonomous vehicles, in smart cities, for example? Or even support for robots adapted to interact with humans?
First, remember that this is a Research and Innovation action (RIA) topic with technology readiness level expected at TRL4-5 at the end of the project, so well below the level of complete systems or prototypes deployed on real-world systems. There are other topics (in the Digital Europe programme) that address deployment of existing technologies. Second, the work programme does not mention any application area. You can choose the mandatory three data spaces for your use cases in sectors/areas that you want (as long as it makes sense and your approach is realistic).

Q: Is a more direct application intended for industrial installations, factories? Or is it intended to create a system that can more easily support autonomous vehicles, in smart cities, for example? Or even support for robots adapted to interact with humans?
The topic calls for development of generic and Advanced AI-enabled Cloud/Edge technologies. You need to demonstrate the generic applicability on a number of use cases/applications. We don’t specify which application domains. This depends on how you choose to prove the proposed technologies work.

Q: Does the Cognitive Computing Continuum call include solutions for specific domains, e.g. healthcare?        
A: No specific domains. You need to demonstrate the generic applicability of the proposed technologies.

Q: How many use cases would you like to see in a project for HORIZON-CL4-2023-DATA-01-04?
Use cases need to demonstrate the generic applicability of the proposed technological solution. It is up to the proposers to decide that.

Questions on HORIZON-CL4-2023-HUMAN-01-02

Q: If I understood correctly, you indicated that the Human-01-02 call text does not explicitly mention areas/domains. Yet, in the WP it says: "Pilots should target technological advances with large scale potential impact ion strategically important sectors with large societal impacts such as healthcare, improved working and/or living conditions, etc." Could you kindly clarify?
They are some examples provided. But even healthcare allows for a variety of use cases. So after all they are examples. What is important is what was also displayed on the slide: All proposals should demonstrate the assessment criteria upon which the proposed sectors/use-cases have been selected (e.g. in terms of socioeconomic factors, etc.).

Questions on HORIZON-CL4-2023-DATA-01-02 

Q: Is it mandatory to have partners from the Common Data Spaces in the consortia? Also where can we find those funded Data Spaces?
Not mandatory, because the work programme states: "Based on an analysis of cross-fertilisation potential of data re-use, the proposal should include use cases or pilots addressing or involving at least three different common European Data spaces and/or related ecosystems. In particular, they should create links with the Data Spaces support centre funded under the Digital Europe programme, and work in close collaboration with the emerging Common European data spaces in order to ensure interoperability and coordination of data architectures. ". But you should create all the necessary liaisons and collaborations: "Proposed actions should build on or seek collaboration with existing projects and develop synergies with other relevant European, national or regional initiatives, funding programmes and platforms."

Q: One of the stated expected outcomes is: “successful deployment of data spaces…”. Should the proposals contribute to the deployment of a new data spaces? or just provide technological support for existing/future data spaces?
The requirement to have use cases that involve at least three different Data Spaces is there because we don’t want the DATA-01-02 to work on artificial or unrealistic problems and use cases, but tackle those issues that are real challenges when setting up and operating data spaces. We do not require that all the three data spaces are already in deployment at the time of submission, that would be unreasonable (as very few of them are yet in existence). But the link to (future) data spaces should be a bit more than hypothetical, and there should be a clear plan and timeline to the use case with a given data space. If in doubt, please consult the Digital Europe work programme, and the Data Spaces Support Centre https://dssc.eu/ to find out which data spaces are being planned and at which stage of development they are.

Q:  Can you please elaborate more on what is meant by 1) “human language issues” in the data life cycle and the 2) “…addressing the social and cultural factors”.
The WP states: "Systems and approaches should be able to process human-generated and human-related data (e.g. speech, text, images) and put data into context (including cultural, linguistic and social context)". Firstly, many of the Data Spaces will contain human language (text, speech), and language understanding as a built-in feature may be necessary, depending on the application. Secondly, for this purpose, one of the Data Spaces being funded under Digital Europe is the Language Data space, containing language data and supporting development of language models that support systems that "understand" human language. Therefore, human language (and other human factors) are an essential element of many data spaces (e.g. health, culture, tourism data spaces).

Questions on HORIZON-CL4-2023-HUMAN-01-01

Q: With respect to call Efficient trustworthy AI (HORIZON-CL4-2023-HUMAN-01-01). It states proposals should address one of the focus areas: one of these areas mentions machine learning methods that address one-shot learning, continuous and/or lifelong learning. We have set up a consortium that focusses on continual learning (and we address transfer learning, few shot learning in relation to continual learning). Would the committee welcome projects which focus on one set of algorithms/settings (namely continual learning) or is this considered too restrictive.
You will have to address (fully) one of the to focus areas. And then it depends upon the proposal to convincingly demonstrate that it addresses the focus.

Q: Shall use cases be oriented to AI functionality rather to specific sectors? (i.e. use cases targeting AI features instead of verticals) How much robotics involvement should proposals have under this topic? Is it expected a use case focused on robotics?
Please read the full topic description. If this is not explicitly mentioned there, it is up to you to convincingly show how the proposed use cases/sectors demonstrate the progress beyond the state of the art in one of the two focus areas of the topic.

In the first instance please read the specific call for which you are applying. 

You can watch the recording of the Info Day and Brokerage event and download all presentations here. 

If you have questions about Adra membership or are interested in joining please find more information here.