Grants8 | The most comprehensive grants search on the web

Grant Analysis

Purpose & Target

Grant Purpose and Target

Core Objective: To develop robust and trustworthy Generative AI solutions specifically for robotics and industrial automation, aiming to boost productivity and competitiveness within the European manufacturing sector.
Target Recipient Type and Size: Primarily organizations engaged in advanced Artificial Intelligence (AI) research and industrial application. This includes businesses of various sizes, with a focus on those that can collaborate with or support the manufacturing industry, including SMEs and startups as third-party recipients.
Designation: SECTOR-SPECIFIC (Manufacturing, Robotics, Artificial Intelligence).
Geographic Scope: Open to legal entities established in EU Member States, Iceland, Norway, Canada, Israel, the Republic of Korea, New Zealand, Switzerland, and the United Kingdom.
Key Filtering Criteria: Focus on Generative AI, robotics, industrial automation, and projects that contribute to the EU's strategic autonomy in digital technologies.
Grant Frequency and Program Context: This is a single-stage call within the Horizon Europe Work Programme 2025, specifically under 'Digital, Industry and Space'. It is part of the broader GenAI4EU initiative and aligns with the AI, Data, and Robotics (ADRA) and Made in Europe European Partnerships.

Financial Structure

Total Budget for this Topic: EUR 85,000,000.
Minimum Funding per Project: EUR 40,000,000.
Maximum Funding per Project: EUR 45,000,000.
Currency: EUR.
Mandatory Financial Support to Third Parties (FSTP):
- A minimum of EUR 10,000,000 from the proposal budget must be allocated to FSTP.
- The maximum amount for each single third party can reach up to EUR 500,000 (exceeding the usual EUR 60,000 limit).
- The maximum amount for each action implemented by a third party or a consortium of third parties is up to EUR 2,000,000.
- FSTP is specifically for the fine-tuning phase of Generative AI applications tailored to industry use cases.
Foundation Model Development Budget: Up to EUR 30,000,000 of the proposal budget should be dedicated to the development of the foundation model.
Eligible/Ineligible Costs: Details on eligible and ineligible costs are described in Annex G of the Work Programme General Annexes.
Computing Resources: If significant computing resources are needed for pretraining, applicants must convincingly describe their strategy to access external High-Performance Computing (HPC) facilities, such as EuroHPC or national centers.

Eligibility Requirements

Organizational Type & Legal Status

Eligible: Legal entities established in any EU Member State, Iceland, Norway, Canada, Israel, the Republic of Korea, New Zealand, Switzerland, and the United Kingdom.
Exclusions for Strategic Reasons: Entities based in an eligible country but controlled, directly or indirectly, by a non-eligible country or entity are generally excluded. Participation is possible only if they provide positively assessed guarantees demonstrating no negative impact on the Union's strategic assets, interests, autonomy, or security.
Specific Exclusion: Entities identified as high-risk suppliers of mobile network communication equipment (or those fully/partially owned or controlled by such suppliers) cannot participate.

Financial & Operational Capacity

All applicants must demonstrate sufficient financial and operational capacity, as detailed in Annex C of the Horizon Europe Work Programme General Annexes.
Beneficiaries awarded funding must be able to provide 'Financial Support to Third Parties' (FSTP) as part of their project implementation.

Project Scope & Focus

Proposals must clearly identify if they are addressing 'Type A' (Generative AI for Robotics for industrial automation) or 'Type B' (Trustworthy and robust generative AI for improved manufacturing) outcomes.
The research must lead to the development of general-purpose AI models and tools, not just narrowly focused solutions.
Projects are expected to align with European values, principles, and regulatory frameworks, particularly the AI Act.

Consortium Requirements

Consortium approach is expected, particularly involving major manufacturing industry players for Type A projects, and user industries for providing data and use-cases.
For FSTP, third parties can be single companies (including SMEs/Start-ups) or small consortia (e.g., user industry with 1-2 AI developer/integrator partners).

Application Process

Application Deadline: 2025-10-02T00:00:00+0000.
Application Submission: Proposals must be submitted via the designated Submission System on the Funding & Tenders Portal.
Submission Model: This call uses a single-stage submission procedure.

Required Documentation & Materials

Application Form: Use the standard application form specific to this call (HE RIA, IA) available in the Submission System.
Proposal Content: Proposals must include a detailed business case and an exploitation strategy. For these specific sections, the standard page limit in Part B of the application form is exceptionally extended by 3 pages.

Evaluation & Selection Process

Evaluation Criteria: Award criteria, scoring systems, and thresholds are detailed in Annex D of the Work Programme General Annexes.
Evaluation & Submission Procedures: Processes for submission and evaluation are described in Annex F of the Work Programme General Annexes and the Online Manual.
Timeline: An indicative timeline for the evaluation process and grant agreement preparation is provided in Annex F of the Work Programme General Annexes.

Post-Award & Reporting Obligations

Financial Support to Third Parties (FSTP): If awarded, beneficiaries are legally required to provide FSTP as part of their project. This mechanism, including its legal and financial setup, is governed by Annex G of the Work Programme General Annexes.
Progress Tracking: Projects are expected to track progress using quantitative Key Performance Indicators (KPIs) to demonstrate the impact of Generative AI solutions.
Result Dissemination: Communicable project results should be shared with the wider European R&D community, utilizing platforms like the AI-on-demand platform and other relevant digital resource platforms.

Evaluation Criteria

Award Criteria & Portfolio Balance

Proposals will be assessed against criteria, scoring, and thresholds outlined in Annex D of the Work Programme General Annexes.
To ensure a diverse portfolio, at least one highest-ranked proposal from 'Type A' and one from 'Type B' will be selected for funding, provided they meet all minimum thresholds.

Expected Impact & Strategic Relevance

Type A Projects: Must demonstrate contributions to developing advanced foundation models for robotics, enhancing autonomy and generalization, and validating these through industrial automation use-cases.
Type B Projects: Must show contributions to increased productivity, improved product/process certification and compliance, enhanced reliability, efficiency, and sustainability of manufacturing processes, and facilitated installation/commissioning/decommissioning of production facilities.
All proposals must convincingly demonstrate their expected positive impact on the competitiveness of the chosen application sector.
Proposals should contribute to creating and disseminating general-purpose AI models and tools that align with European values, principles, and the AI Act.

Technical Excellence & Methodology

Type A Specifics: Emphasis on integrating 'Human-in-the-loop' mechanisms, addressing data volume challenges for training (e.g., via model distillation for edge portability), and ensuring training data is diverse to mitigate bias.
Type B Specifics: Projects should exploit advanced Generative AI models (e.g., foundation models, LLMs, transformers, multimodal AI) with a focus on manufacturing data. Methodologies should address robustness, trustworthiness, data management, automation, and tools for rapid deployment and innovation cycles.
Research activities should explore training methodologies for foundation models, particularly their ability to process multimodal data for robotic decision-making.
Rigorous testing, including simulation and physical experiments, is required for model validation.

Strategic Alignment & Collaboration

Projects must implement the co-programmed European Partnerships on AI, Data, and Robotic (ADRA) and Made in Europe.
Proposals should include tasks for cohesion activities with ADRA and the 'GenAI4EU central Hub' (HORIZON-CL4-2025-03-HUMAN-18).
Synergy and collaboration with existing projects, and other relevant international, European, national, or regional initiatives are highly valued.
Proposals should leverage and contribute to tools and resources available on the AI on demand platform, ensuring knowledge sharing within the European R&D community.

Business Case & Exploitation

A detailed business case and exploitation strategy are mandatory parts of the proposal, for which additional page limits are provided.
The selection of use cases must be justified by their significant business dimension, ensuring sufficient resources are allocated to achieve impactful outcomes.

Compliance & Special Requirements

Regulatory Compliance & Ethical Standards

AI Act Compliance: Projects must develop models that align with European values, principles, and comply with the AI Act requirements, ensuring trustworthiness and reliability.
Safety & Cybersecurity: Proposals must explicitly address both the physical safety of robotic operations and robust cybersecurity measures to protect against digital threats and maintain system integrity.
Bias Mitigation: Training data, especially for human interaction contexts, should include diverse individual characteristics (e.g., gender, age, racial/ethnic background) to actively mitigate potential bias and discrimination.

Strategic Autonomy & Security Considerations

Technological Independence: Projects must avoid creating technological dependency on non-EU sources, helping to safeguard the Union's strategic assets, interests, autonomy, and security.
Controlled Entities: Legal entities established in eligible countries but controlled by non-eligible countries/entities must provide guarantees to ensure their participation does not compromise EU strategic interests. These guarantees must confirm control over the applicant entity, restrict access by non-eligible entities to sensitive information (potentially requiring national security clearance for personnel), and ensure IP ownership remains within the recipient and is not exported without approval.

Data Management & Intellectual Property (IP)

Industrial Data Handling: Consortia, particularly those involving manufacturers, must define upfront mechanisms for pooling and collectively providing sufficiently large datasets for model training, potentially using trusted third parties to handle sensitive industrial data.
IP Ownership: All intellectual property and results generated by the action must remain within the recipient organization during and after the project. This IP must not be subject to control or restrictions by non-eligible countries/entities, nor can access to it be granted from outside eligible countries without prior approval from the country of establishment.
Data Interoperability: Where relevant, proposals should address interoperability for data sharing, prioritizing open specifications and standards to enable effective cross-domain data communities.

Collaboration & Dissemination Expectations

European Partnerships: Projects are expected to implement and contribute to the co-programmed European Partnerships on AI, Data, and Robotic (ADRA) and Made in Europe.
Synergies: Proposals should build upon or seek collaboration with existing and future projects, fostering synergies with other relevant international, European, national, or regional initiatives.
Community Engagement: Applicants are encouraged to leverage tools available on the AI on demand platform and disseminate communicable results to the European R&D community through this platform and other relevant digital resource platforms.

Grant Details