Loading...
JARVIS
1 phase
Strategic Analysis
JARVIS OC2 Co-Development (Stream 1) is a cascade Open Call under the Horizon Europe Programme, run by the JARVIS consortium alongside the parallel External Pilots track. It invests a total of €800,000 to enhance the four core JARVIS industrial pilots.
Applicants develop independent, ROS2-deployable add-on modules (software modules, perception systems, interaction technologies, mechatronic components) that interface with the existing JARVIS architecture and extend the capabilities of one of the four pilots.
Develop an innovative ROS2-deployable HRI/AI module addressing one predefined pilot Topic
Integrate and validate the module within the selected JARVIS pilot testbed (TRL6)
Implement ROS2-compatible communication interfaces with the JARVIS architecture
Participate in user-centric evaluation within the JARVIS pilot
Demonstrate industrial relevance and a credible exploitation pathway
Address security, privacy and safety
Follow the structured 3-sprint development process with topic-specific KPIs
Value-added ROS2-deployable module integrated and validated at TRL6 in a JARVIS pilot
Demonstrated module performance against topic-specific KPIs
Bidirectional knowledge exchange with the JARVIS consortium and EU HRI ecosystem
Credible industrial adoption / commercialization pathway
Expanded JARVIS ecosystem and accelerated adoption of human-centric industrial HRI
Horizon Europe Programme
highThe evaluator will assess how well the proposed project aligns with the overall objectives of the Horizon Europe Programme, particularly in fostering research and innovation with significant industrial and societal impact.
Commission Recommendation 2003/361/EC
highThe evaluator will verify that the applicant SME strictly adheres to the EU definition of Micro, Small, and Medium-sized Enterprises as outlined in this Recommendation, ensuring eligibility for funding.
Commission Regulation No 651/2014, art. 2.18
highThe evaluator will check that the applicant SME is not considered an 'enterprise under difficulty' according to Article 2.18 of this Regulation, which would render them ineligible for funding.
General Data Protection Regulation (EU) 2016/679 (GDPR)
highThe evaluator will assess the proposal's commitment to and methods for ensuring full compliance with the GDPR, particularly regarding data collection, processing, anonymization, and user consent in human-robot interaction scenarios.
EU restrictive measures under Article 29 of the Treaty on the European Union (TEU) and Article 215 of the Treaty on the Functioning of the European Union (TFEU)
highThe evaluator will verify that no applying entity is subject to EU restrictive measures (sanctions) as defined by these articles, which would render them ineligible for funding.
Horizon Europe Regulation 2021/695
mediumThe evaluator will confirm that all participating entities are legally established in an EU Member State or a Horizon Europe Associated Country as governed by this Regulation.
Council Regulation (EC, Euratom) No 1605/2002, Article 111
mediumThe evaluator will ensure that there is no risk of double funding for the same activity from the EU budget, in compliance with the fundamental principle outlined in Article 111 of this Financial Regulation.
European Code of Conduct for Research Integrity
mediumThe evaluator will check that the proposal addresses ethical issues and adheres to the fundamental ethical principles outlined in this Code of Conduct, including submitting a self-assessment ethics questionnaire.
Council Implementing Decision (EU) 2022/2506
lowThe evaluator will consider this Decision in the context of EU restrictive measures, particularly if the applicant's country of establishment is Hungary, to ensure compliance with Union budget protection measures.
To be eligible for JARVIS OC2 Co-Development, applicants must meet the following criteria (Section 2, Section 5.1 Table 1):
Evaluation runs in three phases. Phase 2 is the scored external evaluation against four criteria, each on a 0-5 scale (decimals allowed; per assessment dimension: 1 = clearly satisfied, 0.5 = partial, 0 = inadequate).
| Criterion | Focus | Threshold |
|---|---|---|
| 1. Concept & Innovation | Clear technical problem of the selected Topic; clear module concept; innovation vs state-of-the-art; technical feasibility within the 9-month programme. | ≥3/5 |
| 2. Impact | Applicability beyond the specific pilot; IPR management; European industrial competitiveness; scalability/transfer roadmap; economic & societal impact with measurable KPIs. | ≥3/5 |
| 3. Technology Implementation Approach | Work plan, methodology, resources, JARVIS integration; ROS2 interfaces/APIs; milestones, deliverables, risks; testing/validation in the pilot; path to TRL6. | ≥3/5 |
| 4. Consortium / Applicant Team & Use of Resources | Team fit; demonstrated expertise (papers, videos, results); realistic budget allocation; sufficient technical & organisational capacity. | ≥3/5 |
Overall threshold: ≥15/20. Proposals scoring <3 in any criterion or <15 overall are rejected.
A panel validates the ranking, aiming for up to 2 proposals per pilot to proceed to the Phase 3 interview.
Funding model: lump sum, up to €100,000 per project; total envelope €800,000 for 8 projects.
The grant finances: employee work; software/hardware (depreciation value only); travel for project deployment or JARVIS activities; participation in JARVIS-associated events and promotion. Under the lump-sum model, beneficiaries do not document detailed costs, but the EC may run technical checks and reviews. LMS (project coordinator) is the sole organisation making payments, ~40 working days after a positive review.
The maximum total support a single entity can receive across all JARVIS Open Calls (OC1, OC2, …) is €200,000; OC1 beneficiaries must keep their combined award under this cap.
The funded programme is a 9-month structured process in three sprints. Development happens at the applicant's premises; integration and validation happen within the JARVIS pilot testbeds in collaboration with the pilot partner.
Section 8.1 of Annex 1.1. Background IP remains with its owners; foreground results IP follows Horizon Europe cascade rules. Each Topic specifies source-code availability expectations; modules are delivered in ROS2-deployable form (docker image, executable, or open repository) and integrations must respect the JARVIS architecture's interface terms. A clear IPR / exploitation strategy is assessed under the Impact criterion.
Consortium dimensioning expectations for JARVIS OC2 Co-Development.
Multi-actor — programme-side, not applicant-side. Unlike the External Pilots track, the applicant does not bring a Use Case Provider: the JARVIS pilot is the fixed use case. End-user engagement happens through collaboration with the pilot partner and its operators during requirements (Sprint 1) and integration/validation (Sprint 3), and through user-centric evaluation.
Interdisciplinarity — NOT EXPECTED. The call seeks a focused technology team able to deliver and integrate a ROS2 module, not a multi-disciplinary academic consortium.
Transdisciplinarity — NOT EXPECTED. The work is industrial module development and integration.
SSH Integration — NOT EXPECTED on the applicant side. The JARVIS programme provides a dedicated SSH Mentor (SSM) who ensures market fit, user-centricity and alignment with societal/regulatory considerations.
Consortium design implication: a credible single SME/startup (optionally plus an RTO integrator) with demonstrated robotics/AI capability and a clear plan to integrate into the chosen pilot is the strongest signal — not consortium size.
Required before submission: the F6S application form (Annex 3.1) and the proposal template (Annex 2.1).
Required at contracting (once selected):
After selection for funding:
Each sprint concludes with a remote review (teleconference) where beneficiaries present progress and respond to JARVIS experts. Deliverables are due 15 calendar days before the end of each sprint.
Accepted deliverables release the corresponding lump-sum payment (~40 working days). Rejected deliverables must be revised; repeated unsatisfactory outcomes can lead to termination. A short extension may be granted if Sprint 3 review is unsatisfactory.
Sprint-level technical KPIs (S1.1-S1.4, S2.1-S2.5, S3.1-S3.4) plus the topic-specific KPI (KPI-Topic), and dissemination KPIs (≥2 social-media posts per sprint and ≥150 interactions in total; one JARVIS blog post per sprint).
The call requires you to pick or distribute several structuring elements before submission. These will be locked at proposal creation time and drive the project backbone. Read-only here — the Coach will guide you through them when you create your proposal.
jarvis-consortium-role
Technology Developer (Lead Partner)
Technology Integrator / Knowledge Provider (Optional)
jarvis-codev-topic
COLLINS · Large and deformable parts models for impedance-control co-manipulation
Model large, deformable carbon-fibre composite sheets (analytical or learning-based, e.g. PINNs / Learning-from-Demonstration) and deliver a ROS2 module that adapts robot impedance and trajectories in real time for fluid, precise human-robot co-manipulation in aircraft seat layup.
Pilot: COLLINS (Aeronautics Manufacturing). Target TRL6. Technical mentor: TECNALIA. Testbed: San Sebastián, Spain.
COLLINS · Zero-Interface Human-Robot Interaction in agile manufacturing
Develop a zero-interface module enabling implicit, hands-free HRI in composite layup co-manipulation by sensing and interpreting human neural/muscular/physiological signals (EEG, EMG) for cognitive-state and gesture/intention detection.
Pilot: COLLINS (Aeronautics Manufacturing). Target TRL6. Technical mentor: TECNALIA. Testbed: San Sebastián, Spain.
COLLINS · Real-time robot trajectory and dress-pack collision avoidance
Build a ROS2/MoveIt2-compatible module that verifies robot trajectories online to avoid collisions and self-collisions with dress packs and surrounding geometry, validated on a UR20 robot with pneumatic tubing and 3D-sensor cabling.
Pilot: COLLINS (Aeronautics Manufacturing). Target TRL6. Technical mentor: TECNALIA. Testbed: San Sebastián, Spain.
TOFAS · Intelligent mechatronic system for collaborative manipulation of linear flexible components
Develop an intelligent mechatronic system for safe, stable, compliant human-robot co-manipulation of linear flexible components (cables, tubes) — routing, positioning, alignment and shared handling — during TOFAS battery-pack assembly.
Pilot: TOFAS (Automotive Manufacturing). Target TRL6. Technical mentor: LMS. Testbed: Patras, Greece.
TOFAS · Wearable-augmented human tracking for uncertainty reduction in close-range HRC
Build a wearable-augmented sensing/interpretation module that improves human body tracking and activity recognition under clutter and occlusion, and adds physiometric/cognitive-load indicators for workload- and safety-aware adaptive collaboration in battery-pack assembly.
Pilot: TOFAS (Automotive Manufacturing). Target TRL6. Technical mentor: LMS. Testbed: Patras, Greece.
TOFAS · Digital safety validation and risk-aware design for close-range HRC
Develop a digital safety-validation module (optionally VR/XR-enabled) that uses generative AI to auto-generate and simulate diverse HRC scenarios (layouts, motion paths, task sequences, operator behaviours) and flag safety-critical situations for TOFAS battery-pack assembly.
Pilot: TOFAS (Automotive Manufacturing). Target TRL6. Technical mentor: LMS. Testbed: Patras, Greece.
EDF · Optimization of waste-volume packaging
Develop a module that optimises the packing of nuclear decommissioning waste into certified containers to reduce the number of containers sent to limited-capacity storage facilities, validated with robotic handling in a TRL6 context.
Pilot: EDF (Nuclear Plants Decommissioning). Target TRL6. Technical mentor: CEA. Testbed: Palaiseau, France.
EDF · Anomaly detection for teleoperation and skills execution
Enhance teleoperation monitoring (tools, robotic arm, cameras) to detect anomalies (e.g. shear positioning in pipe-cutting, circular-saw cutting, bin placement) and present AI-enhanced, user-centric visual feedback that lets the teleoperator understand anomalies and system status.
Pilot: EDF (Nuclear Plants Decommissioning). Target TRL6. Technical mentor: CEA. Testbed: Palaiseau, France.
EQUINOR · Robust self-localization and mapping in semi-outdoor process-plant environments
Develop a robust ROS2/Nav2-compatible self-localization and mapping module with multi-sensor fusion (3D lidar, ZED2 stereo) for mobile robots in semi-outdoor, partially GNSS-denied plant environments; optional barrier-chain detection; alignment with CAD/digital maps.
Pilot: EQUINOR (Offshore Inspection & Maintenance). Target TRL6. Technical mentor: SINTEF. Testbed: Trondheim, Norway.
EQUINOR · Online automatic detection of damages and operational features
Develop a vision-based inspection module that detects and categorizes asset damage (corrosion, coating loss, cracks, leaks) and verifies operational features (valve car seals, barrier chains) during robot operation, with clear evidence-supported analysis outputs.
Pilot: EQUINOR (Offshore Inspection & Maintenance). Target TRL6. Technical mentor: SINTEF. Testbed: Trondheim, Norway.
EQUINOR · Autonomous intervention with mobile manipulator using foundation models
Develop an autonomous intervention system for a mobile manipulator (UR5e + gripper + ZED2i, ROS2 Humble) using robotics foundation models (VLA / multimodal / generalist policies), built first in simulation then deployed to perform push-button, switch, cabinet and valve tasks under varying outdoor lighting.
Pilot: EQUINOR (Offshore Inspection & Maintenance). Target TRL6. Technical mentor: SINTEF. Testbed: Trondheim, Norway.
5 key insights you must internalise before writing. Each is grounded in the call text and tells you what evaluators will actually look for. Share these with your consortium before drafting.
Co-development Topics are predefined by the JARVIS pilots. Concept & Innovation (Criterion 1) and tie-break Rule 2 reward proposals that demonstrate deep understanding of the chosen pilot's operational context, baseline technologies and integration interfaces. A generic module not tightly fitted to the selected Topic and pilot will score poorly regardless of its standalone quality.
Source: Annex 1.1 Section 3 (Topics per pilot) + Section 5.2.1 Criterion 1 + Section 5.2.3 Rule 2
The deliverable is a ROS2-deployable module integrated into the JARVIS pilot testbed at TRL6 (Sprint 3), not just a standalone prototype. Technology Implementation (Criterion 3) checks ROS2 interfaces/APIs, milestones, deliverables, risks and a credible path to TRL6 in the 9-month window. De-risk the interface design and integration concept in Sprint 1.
Source: Annex 1.1 Section 1.1 (TRL6) + Section 7.1 (Sprints) + Section 5.2.1 Criterion 3
Payments follow a 20% / 60% / 20% lump-sum schedule — the 60% middle instalment is paid only after the Sprint 2 Review (development & deployment, ~M8). Plan cash-flow and effort around delivering a working, interface-tested module prototype by end of Sprint 2; this is where projects stall.
Source: Annex 1.1 Section 7.1 (payment 20/60/20) + Table 2
Each legal entity may take part in only one proposal across both simultaneous JARVIS Open Calls (Co-development and External Pilots). If an entity appears in multiple submissions, all proposals involving it are automatically declared ineligible at Phase 1.
Source: Annex 1.1 Section 2.1.5.1 (Multiple submission)
Tie-break Rule 3 favours higher impact and market potential. Beyond the technical module, evidence a clear IPR/exploitation strategy and a credible commercialization pathway (Criterion 2). Funding is capped at €100k and the cumulative cap across all JARVIS Open Calls is €200k per entity.
Source: Annex 1.1 Section 5.2.1 Criterion 2 + Section 5.2.3 Rule 3 + Section 2.1.1 (€200k cap)
Phase 2 — full Writing Strategy Matrix (scoring pressure map, keywords to embed, compliance gates, sections priority order) is planned via a dedicated evaluation strategist agent. See Notion PRD "Call Analysis enrichie".
The AI has drafted potential core elements based on the call analysis. To start building your project proposal structure, select the elements that resonate with your consortium's concept. You can refine and rewrite them fully once your project workspace is created.
Each of the 11 predefined Topics expresses a concrete capability gap in a JARVIS pilot — deformable-part modelling, zero-interface HRI, collision avoidance, wearable tracking, safety validation, waste packaging, anomaly detection, SLAM, damage detection, autonomous intervention — that the existing pilot systems cannot yet address.
Promising HRI/robotics modules stall between lab prototype (TRL4-5) and industrial relevance because innovators rarely have access to real, instrumented pilot environments in regulated sectors such as aeronautics, automotive, nuclear and offshore.
Heterogeneous, non-interoperable software limits reuse of validated capabilities. A ROS2-deployable add-on validated inside the JARVIS architecture lowers the barrier to cross-system reuse.
Deploying robots for inspection, maintenance and intervention still demands extensive manual programming, slowing industrial adoption — a gap that advanced perception, cognition and foundation-model approaches can close.
Operators, technicians and engineers at the four JARVIS pilot testbeds (aeronautics, automotive, nuclear decommissioning, offshore) whose tasks the developed module augments — the people who co-define requirements and take part in user-centric validation.
The industrial pilot owners providing the fixed use case and testbed. They adopt and validate the integrated module and own the post-project replication decision.
The primary technical mentors (TECNALIA / LMS / CEA / SINTEF) and the wider JARVIS consortium who integrate the module into the pilot architecture and exchange technical insight bidirectionally.
SMEs and startups providing HRI, AI and robotics components — potential adopters and replicators of the validated module beyond the JARVIS pilots, reached through dissemination and exploitation activities.
The broader European Human-Robot Interaction community whose architecture and practice benefit from the knowledge exchange generated by the co-development projects.
Design and implement an innovative HRI/AI/perception/mechatronic add-on for the selected Topic, delivered in ROS2-deployable form (docker image, executable or open repository) compatible with the JARVIS architecture.
Move from TRL4-5 prototype to a module integrated into the JARVIS pilot environment and validated in collaboration with the pilot partner at target TRL6 (Sprint 3).
Finalise topic-specific KPIs in Sprint 1, then evidence module performance against them through testing and user studies within the pilot, addressing usability, operator acceptance and effectiveness.
Produce an exploitation/commercialization roadmap and contribute to bidirectional knowledge exchange with the JARVIS consortium and the European HRI ecosystem.
Each validated add-on extends the four JARVIS pilots with new HRI capabilities, raising the capability and reuse value of the JARVIS architecture across sectors.
Demonstrating modules at TRL6 in real industrial pilots reduces perceived risk for adopters and accelerates uptake of human-centric robotics in aeronautics, automotive, nuclear and offshore.
Funding, mentoring and testbed access help SMEs/startups mature and exploit HRI technologies, strengthening European competitiveness in industrial robotics.
Human-centric HRI modules reduce strain, cognitive load and exposure to hazardous conditions, improving occupational safety and quality of work for operators.
Work packages / sprints (3)
Sprint 1: Requirements
M1–M2 · 20% payment
Sprint 2: Development & Deployment
M3–M7 · 60% payment
Sprint 3: Integration & Validation
M8–M9 · 20% payment
Key results / TRL targets (1)
TRL6 demonstrator — module integrated and validated within the JARVIS pilot testbed
TRL 4 → TRL 6
Deliverables (8) — imposed by the call
Sprint 1 Report (system architecture, integration approach, technical specs, topic-specific KPIs, roadmap)
M2One-pager dissemination material
M2Sprint 2 Report (implemented functionalities, progress, next integration steps)
M7Video demonstration of the module prototype in a relevant environment
M7Publishable summary of progress (incl. ≥2 social-media posts)
M7Final Report (integration outcomes, KPI evaluation, exploitation roadmap)
M9Video demonstration of the integrated module within the JARVIS pilot testbed (TRL6)
M9Publishable summary of achievements
M9Milestones (3) — imposed by the call
Sprint 1 Review (requirements approved)
M320%Sprint 2 Review (module developed & tested)
M860%Sprint 3 Final Review (TRL6 integration validated)
M1020%Fully managed by GrantForge
Click Create proposal above to start writing this form inside GrantForge. The structure below is what we generate for you.
Applicants complete the official Annex 2.1 JARVIS OC2 Proposal Template (Co-Development) (DOCX) and submit it on the F6S platform. Page limits and formatting rules follow the guidelines (Annex 1.1). The proposal template becomes the Implementation Plan annex to the Sub-Grant Agreement.
Annex 3.1 JARVIS OC2 F6S Application Form (Co-Development) is provided as an offline support document; the actual F6S online form must be completed at submission time.
No separate Part C budget spreadsheet. Budget is captured within the F6S application form under the lump-sum structure (20% / 60% / 20% across the three sprints), capped at €100,000 per project.
Canonical EU Funding & Tenders Portal page for this call.
https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/competitive-calls-cs/13733
Institutional / marketing website of the parent programme.
https://jarvis-project.eu/stream-1-co-developers/
Where you submit your proposal (F6S, opencalls.fund, EPSS, MS Forms…).
https://www.f6s.com/jarvis-oc2-co-development/apply
Official documents pack (applicant guide, templates, annexes).
https://www.jarvis-project.eu/open-calls/open-call-2/