About the Open Call

Launch date: 9 November 2026, 09.00 CET.

Target applicants:
SMEs, startups, mid-caps, large companies, or Research and Technology Organisations (RTOs) — acting as data providers and/or developers of data-collection tools (hardware and/or software), to receive funding and support. Applicants must be legal entities established in an EU Member State, Iceland, Norway, or an eligible associated country (Canada, Israel, Republic of Korea, New Zealand, Switzerland, United Kingdom)

Scope of activities: The core mission of the call is to expand and democratise the data foundations that train the GRAIL Foundational Robotics Model (GFRM) — the first European general-purpose, agentic Mixture-of-Experts architecture unifying Generative AI and Robotics. Activities must fall directly within the scope of GRAIL Open Call, which concerns the development of technical solutions focusing on data collection, crowdsourcing, low-cost capture devices, and open-benchmarking tools to resolve the "Robotic Data Gap".

Eligible activities
Eligible activities must focus on creating or improving the data assets that will feed the GRAIL foundation model. The indicative activities that qualify for financial support are the following:
Generalist Data Foundations: Creation of broad datasets across robots and tasks, capturing primitive skills (grasping, pushing, insertion, handover) with multimodal sensors. Emphasis on lab-scale but diverse conditions, including rigid, deformable, and cluttered interactions, to serve as transferable baselines for later UCs.
Democratizing Data Capture. Low-cost, plug-and-play kits for data collection, with multimodal sensors easily retrofittable to lab robots. Citizen-driven contributions via phone apps and gamified platforms, enabling non-roboticists to record demonstrations and interactions, increasing diversity and accessibility of training data.
Lab-scale Generalization Stress Tests. Benchmarks to evaluate zero-/few-shot generalization, sim-to-real transfer, and robustness under varying lighting, clutter, and human presence. Micro-scenarios for safe HRC (handover, collaborative navigation) tested in controlled lab conditions.
Open Benchmarking Tools. Shared tests, evaluation boards, and baseline models that allow consistent measurement of performance across partners and future UCs. Focus on interpretability, comparability, and transparent reporting of progress.

We are funding the following topics:
Topic 1: Dexterous and precise tasks with robotic hands
Topic 2: Long-horizon stationary robot tasks
Topic 3: Long-horizon mobile robot tasks
Topic 4: Industrial Robot Tool-Use Tasks
Topic 5: Egocentric Video Data Collection for Dexterous Tasks
Topic 6: Low-Cost Multimodal Capture for Contact-Rich Manipulation
Topic 7: Plan–Fail–Repair: Gamified Collection of Robot Reasoning Data
Topic 8: Open generalization-stress benchmarking kit: physical stress artefacts, sweep protocol and a portable evaluation harness with reliability and human-in-the-loop reporting

Funding available: Financial Support to Third Parties (FSTP) up to 100,000 EUR per project (Lump Sum).

Why participate? The Genesis Open Call will select at least 10 individual third-party projects, with each selected entity receiving a fixed lump sum of up to €100,000, delivered over a maximum of 8 months across three stages:
Stage 1. Individual Mentoring Plan (IMP) — 1 month — up to €23,000
Stage 2. Technology design and development, and Data acquisition — 4 months — up to €50,000
Stage 3. GRAIL Data and tech integration — 3 months — up to €27,000
Successful applicants will also be enrolled in the GRAIL Mentoring Programme and supported by GRAIL's technical coaches.