Loading…
Strategic Analysis
BattleVerse funds six-month experiments by SMEs and start-ups that add AI and simulation capabilities to its military mission-planning ecosystem, in one of three topics: LLM-based decision support, multi-physics multi-domain simulation, or distributed AI for coordinated UxV swarms. Proposals win on Impact, which has the highest threshold (4 out of 5) and breaks ties: evaluators look for measurable operational KPIs, a credible integration path into BattleVerse and a realistic route to TRL5 and to exploitation in the European defence market.
Everything the call asks for, seen from the call's point of view. Each line shows what answers it, and which partner carries it.
This matrix lists everything the call asks for: outcomes, impacts, scope, the requirements buried in the call text, and policy alignment. Sign up free and GrantForge tracks each line against the concept you build.
| Requirement | Covered by | Carried | Status |
|---|---|---|---|
| Scope activities | |||
| SC01Trustworthy, human-in-the-loop LLM decision supportEnsure AI-generated outputs remain interpretable, decisions explainable and human operators retain full control, addressing explainability, robustness, transparency, accountability, safety and security in line with European principles for human-centred AI. | · | · | Sign up to track |
| SC02LLMs and agentic AI for strategic decision support in multi-domain scenariosUse of LLMs and agentic AI for strategic decision-support in multi-domain operational scenarios, including complex effects, multi-step planning and decision-support tasks. | · | · | Sign up to track |
| SC03On-premises LLMs for operational security and data sovereigntyExperimentation with on-premises LLMs to ensure operational security and data sovereignty in air-gapped or communication-denied environments. | · | · | Sign up to track |
| SC04Integration with MSaaS-compliant simulation environmentsIntegration with MSaaS-compliant simulation environments, including the use of standardised APIs and data exchange formats. | · | · | Sign up to track |
| SC05Symbolic reasoning, knowledge bases and retrieval-augmented approaches with LLMsIntegration of symbolic reasoning, knowledge bases and retrieval-augmented approaches with LLMs to support data-driven operational decision-making. | · | · | Sign up to track |
| SC06Game-theoretic modelling and AI-assisted scenario generation for wargamingGame-theoretic modelling, simulation, automated planning and AI-assisted generation of operational scenarios for mission preparation and training within a wargaming setting. | · | · | Sign up to track |
| SC07Planning under uncertaintyPlanning under uncertainty and dynamic operational conditions. | · | · | Sign up to track |
| SC08Order generation and explanation in natural languageOrder generation and explanation in natural language, including commanding intention and rationale with respect to the expected effects, the opposing forces, potential collateral damages and neutral elements. | · | · | Sign up to track |
| SC09Explainability, traceability and human approval of AI-supported decisionsExplainability, traceability and visualisation of AI-supported decisions, including human approval workflows for AI-generated actions. | · | · | Sign up to track |
| SC10Safeguards against hallucinated or unsafe recommendationsSafeguards against hallucinated, unsupported or operationally unsafe recommendations. | · | · | Sign up to track |
| SC11Surface and submarine warfare simulationSimulation of surface and submarine warfare scenarios, including interactions between naval platforms and underwater systems. | · | · | Sign up to track |
| SC12Multi-physics and multi-scale models of platform mobilitySimulation of physical laws related to platform mobility in air, ground and naval domains, including multi-physics and multi-scale models; Physics-Informed Neural Networks or other surrogate models are possible. | · | · | Sign up to track |
| SC13Modelling of physical and operational effectsModelling of physical and operational effects, including environmental conditions, sensing, communication, electronic warfare or cyber-related interactions. | · | · | Sign up to track |
| SC14Physical environment modelsIntegration of physical environment models (e.g. terrain, atmosphere, electromagnetic spectrum, underwater environments). | · | · | Sign up to track |
| SC15Interactive agents and behavioural modelsDevelopment of interactive agents and behavioural models representing platforms, systems or operational entities. | · | · | Sign up to track |
| SC16Synchronisation and interoperability with Digital Twins and federated simulation servicesSynchronisation and interoperability between simulation components, Digital Twins, sensors, communication systems or federated simulation services. | · | · | Sign up to track |
| SC17Dynamic, adaptive and scalable simulation frameworksSimulation frameworks supporting dynamic, adaptive and scalable operational scenarios. | · | · | Sign up to track |
| SC18Simulation assumptions on distributed situation awarenessProposals shall provide simulation assumptions on the sharing of distributed situation awareness between agents, for example ranging from asynchronous local views to a global uniform view. | · | · | Sign up to track |
| SC19Automatic planning for mission preparation, control and contingenciesOffline deliberative automatic planning for mission preparation; just-in-time planning for mission control and contingencies, including temporal constraints, temporal uncertainties and onboard resource usage; online planning of UxV behaviour and its execution control. | · | · | Sign up to track |
| SC20AI-driven swarm coordination across heterogeneous UxVsAI-driven swarm coordination and collaborative task execution across heterogeneous UxV systems, such as multi-agent reinforcement learning, Markov decision processes or Bayesian methods. | · | · | Sign up to track |
| SC21Distributed AI for communication, shared situation awareness and coordinationDistributed AI approaches supporting communication, shared situation awareness, decision-making and coordination between autonomous platforms. | · | · | Sign up to track |
| SC22Human supervision and control of autonomous systemsHuman supervision and control mechanisms, including human-in-the-loop or human-on-the-loop approaches, task delegation and mixed-initiative decision making. | · | · | Sign up to track |
| SC23ML pipelines for situation awareness and autonomous behavioursDevelopment of machine learning pipelines for training, testing and evaluating situation awareness and autonomous system behaviours. | · | · | Sign up to track |
| SC24MLaaS/MSaaS platforms for model development and validationUse of Machine Learning as a Service (MLaaS/MSaaS) platforms for model development, training and validation. | · | · | Sign up to track |
| SC25Operation in degraded, contested or GPS-denied environmentsOperation of UxVs in degraded communication, contested, GPS-denied or cyber-compromised environments. | · | · | Sign up to track |
| SC26Operations against hostile agents in uncertain, constrained environmentsOperations in opposing, uncertain and strongly constrained environments, including simulation of hostile agents. | · | · | Sign up to track |
| SC27Interoperability between UxVs, sensors, communication nodes and BattleVerse simulation servicesSynchronisation and interoperability of activities between UxVs, sensors, communication nodes and BattleVerse simulation services. | · | · | Sign up to track |
| Expected outcomes | |||
| EO01LLM and multi-agent AI prototypes for planning and decision supportPrototype tools and demonstrators using LLMs and multi-agent AI for planning and decision support, including BattleVerse integration and on-premises deployment, at strategic, operative or tactical level. | · | · | Sign up to track |
| EO02Human-in-the-loop oversight of AI-assisted decision-makingHuman-in-the-loop approaches ensuring oversight, intervention, auditability and traceability in AI-assisted decision-making. | · | · | Sign up to track |
| EO03Explainable, robust and safe AI outputs under uncertaintyMethods improving the explainability, transparency, robustness and safety of AI-generated outputs in uncertain conditions. | · | · | Sign up to track |
| EO04LLMs integrated with symbolic reasoning, simulations, game theory and RAGIntegration of LLM capabilities with symbolic reasoning, simulations, game-theoretic models and RAG-based data pipelines. | · | · | Sign up to track |
| EO05Performance evaluation of AI decision-support systemsPerformance evaluation of AI-enabled decision-support systems. | · | · | Sign up to track |
| EO06Operational insights on LLM decision support in defenceOperational insights on the role, performance and limitations of LLM-based decision-support tools in defence contexts. | · | · | Sign up to track |
| EO07Benchmark synthesis disseminated to the communitySynthesis of experimental results on benchmarks that can be disseminated within the community. | · | · | Sign up to track |
| EO08Prototype multi-domain simulation components or toolsPrototype simulation components or tools supporting modelling of complex multi-domain operational environments. | · | · | Sign up to track |
| EO09Demonstrations of surface, underwater, electronic warfare or cross-domain simulationDemonstrations of simulation capabilities addressing surface, underwater, electronic warfare or cross-domain operational scenarios. | · | · | Sign up to track |
| EO10Improved modelling of physical and environmental factorsImproved modelling of physical or environmental factors affecting operational dynamics in simulation environments. | · | · | Sign up to track |
| EO11Simulation agents or behavioural modelsSimulation agents or behavioural models representing platforms, systems or operational actors. | · | · | Sign up to track |
| EO12Executable simulation components with documented interfacesExecutable simulation components or model packages with documented interfaces, inputs and outputs. | · | · | Sign up to track |
| EO13Validated operational relevance and interoperabilityExperimentation, evaluation and validation of simulation approaches demonstrating operational relevance and interoperability within defence-oriented environments. | · | · | Sign up to track |
| EO14Operational insights on multi-physics simulationOperational insights on the applicability of multi-physics simulation tools for mission planning, training or scenario analysis. | · | · | Sign up to track |
| EO15Prototypes for multi-UxV mission planning and coordinationPrototype solutions or demonstrators supporting mission planning and coordination of multiple UxV platforms. | · | · | Sign up to track |
| EO16Swarm coordination strategies based on ML or distributed AIDemonstrations of swarm coordination strategies based on machine learning or distributed AI approaches. | · | · | Sign up to track |
| EO17Validated human supervision of swarm operationsExperimentation, evaluation and validation of approaches for human supervision and interaction in autonomous swarm operations, to evaluate human appropriation and involvement. | · | · | Sign up to track |
| EO18AI pipelines for training and evaluating coordinated behavioursAI pipelines or models supporting training and evaluation of coordinated autonomous behaviours. | · | · | Sign up to track |
| EO19Resilient autonomy under degraded or contested communicationsDemonstrations of resilient autonomous operation under degraded or contested communication conditions. | · | · | Sign up to track |
| EO20Operational insights on swarm-based autonomous systemsOperational insights on the applicability and limitations of swarm-based autonomous systems in defence scenarios. | · | · | Sign up to track |
| Other requirements | |||
| REQ1Reach at least TRL5 within the six-month experimentProposals define and demonstrate the capacity of the experiment to reach at least TRL5 by the end of the experimental journey, validated in an application-relevant environment. | · | · | Sign up to track |
| REQ2High-level integration pathway with BattleVerseProposals identify their high-level integration approach with BattleVerse (HLA federate or BattleVerse service API client), since detailed interface specifications are only provided at onboarding. | · | · | Sign up to track |
| REQ3Own testing and validation environmentApplicants provide their own environment for testing and validating the solution; Topic 3 validation in an external realistic simulation environment is required. | · | · | Sign up to track |
| REQ4Experiment data processed only in eligible countriesNon-classified experiment data may not be routed, processed or stored through cloud computing or AI services hosted outside the EU, Norway or Ukraine; the technical approach demonstrates compliance with the EDF infrastructure requirements. | · | · | Sign up to track |
| REQ5Participation in the mandatory events, in-person sessions budgetedThe team demonstrates the capacity and resources to attend the kick-off (M1), the Learning Expedition in Paris (M2), the monitoring meetings (M3, M5) and the Final Evaluation and Hackathon in Athens (M6), and budgets the in-person sessions. | · | · | Sign up to track |
| REQ6Dissemination and participation in sector eventsDissemination and exploitation activities during and after the experiment (social media, newsletters, technical publications) and presentation of results and prototypes at sector-specific events. | · | · | Sign up to track |
| Expected impacts | |||
| EI01European strategic autonomy and competitiveness in defenceEnhance European strategic autonomy and competitiveness, and Europe's defence autonomy and technological leadership in complex multi-domain operations (Impact criterion; BattleVerse objectives). | · | · | Sign up to track |
| EI02New SMEs and start-ups in the European defence sectorSupport the growth of SMEs and start-ups in military simulation and facilitate the access of new SMEs into the defence sector, including companies coming from civilian markets. | · | · | Sign up to track |
| EI03Enhanced mission effectiveness through European simulation add-onsBroaden the repository of European military technologies with innovative add-ons that enhance mission effectiveness and realism in simulated scenarios, addressing operational needs and gaps in current Member States' defence solutions. | · | · | Sign up to track |
| Underlying policies | |||
| POL1European Defence Fund RegulationEstablishes the European Defence Fund and its participation rules: eligible countries, ownership and control of participants (Article 9), definition of control (Article 2(6)) and security responsibilities. | · | · | Sign up to track |
| POL2SME definition (Recommendation 2003/361/EC)Defines micro, small and medium-sized enterprises by staff headcount, turnover and balance sheet, including linked and partner enterprises. | · | · | Sign up to track |
The binding rules of this call. Items marked auto are verified by GrantForge from the call and the template. The others are yours to confirm.
At least one SME or start-up
Every experiment includes at least one SME or start-up (Commission Recommendation 2003/361/EC), alone or with up to two complementary partners.
Non-SME partners only when justified
RTOs, academic institutions or other non-SME entities are accepted only where no suitable SME can provide the capability, and the proposal justifies their involvement.
Eligible countries and control
Entities established in an EU Member State, Norway or Ukraine, not controlled from another country (Article 9 EDF Regulation).
Eligible countries
All applicants are legal entities established in an EU Member State, Norway or Ukraine.
Ownership and control
No applicant may be controlled by a legal entity or authority from another country (Article 9 EDF Regulation); a Member State guarantee may allow an entity controlled from a non-associated country (Article 9(4)); those guarantees must address, among other things, independence in carrying out the action, protection of sensitive/classified information, ownership of results/IP and restrictions on third-country access. Applicants must accurately disclose the relevant information in their ownership and control declarations. Infrastructure, assets and executive management are located in an eligible country for the whole duration.
Legal personality
Natural persons are not eligible; applicants are incorporated (with a VAT number) before the deadline. Sole traders without separate legal personality may apply if they meet the other requirements.
One selected experiment per consortium and per organisation
A consortium may apply to several topics but only one of its experiments is selected; an organisation cannot take part in two selected proposals.
Funding ceiling per beneficiary
Up to EUR 60,000 per beneficiary, EUR 180,000 per experiment.
What the call requires for your dossier to be receivable: a mandatory section, a declaration, a document to attach. Nothing here can be checked automatically — confirm each one yourself once it's done.
Declaration of Honour and Ownership Control
Mandatory annex, signed by the authorised representative of each organisation, uploaded on the EMS with the proposal.
Compliance self-assessment table completed
Yes/No statements on acceptance of the Guide for Applicants, GDPR compliance, absence of conflict of interest, absence of non-EU control and consent to future contact.
Justification of any non-SME partner
Where an RTO, academic institution or other non-SME participates, the proposal describes the capability required and explains why no suitable SME can provide it (section 3.3).
No classified information in the proposal
The proposal contains no classified or restricted information; any need to access classified information or security-restricted assets is indicated with a brief justification.
English, official template, 15-page limit
Proposal written in English on the official template, sections 1 to 3 within 15 pages, uploaded as a PDF of 20 MB at most, with a summary of about 300 words on the EMS.
The call requires you to pick or distribute several structuring elements before submission. These are locked at proposal creation time and drive the project backbone. Read-only here.
battleverse-oc1-topic
Topic 1: LLMs for strategic decision-making, symbolic reasoning, game theory, and mission planning under uncertainty
Topic 2: Multi-physics simulation capabilities for surface/submarine warfare, electronic warfare, and continuous combat across land, air, or naval platforms
Topic 3: Distributed AI for autonomous systems (UxVs) with for coordinated swarm operations
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.
Impact needs 4 out of 5, against 3 for Technical Excellence and Implementation, and it is the first tie-breaker. Give it the 5 indicative pages it is allotted: quantified technological and operational KPIs, a concrete contribution to the BattleVerse Innovation Hub, replicability by other defence SMEs and a credible exploitation model.
Source: Guide for Applicants §5.4.2-5.4.3, Tables 5 and 7
TRL5 is the minimum at M6, validated in an application-relevant environment that the applicant provides. Starting from an existing asset at TRL 3 or 4 makes the path credible; state the start and end TRL in section 1.5 and show how D2 evidences the validation.
Source: Guide §2.2; FAQ 4.7, 4.8, 4.12
No interface specification exists before submission. Name the integration route at the right level: an HLA federate (IEEE 1516, Pitch pRTI, RPR-FOM based object model) for simulation components, or a BattleVerse service API client for AI and decision-support components, and plan the integration plan and first connectivity test in the first two months.
Source: FAQ 4.1, 4.2, 4.6
Experiment data may not go through cloud or AI services hosted outside the eligible countries, even when unclassified. A hosted commercial LLM API outside these countries is therefore excluded; Topic 1 explicitly values on-premises LLMs. Say where models run and data is stored in sections 1.4 and 3.5.
Source: FAQ 4.4, 7.5; Declaration of Honour, no-control declaration point 3
RTOs and universities are only accepted when no suitable SME can provide the capability. If one joins, section 3.3 must name the capability and explain why no SME could bring it; budget the in-person events in Paris and Athens for every partner.
Source: Guide §3.1-3.2, §4; Proposal template §3.3
Specific requirements
At least one SME or start-up
Every experiment includes at least one SME or start-up (Commission Recommendation 2003/361/EC), alone or with up to two complementary partners.
Non-SME partners only when justified
RTOs, academic institutions or other non-SME entities are accepted only where no suitable SME can provide the capability, and the proposal justifies their involvement.
Eligible countries and control
Entities established in an EU Member State, Norway or Ukraine, not controlled from another country (Article 9 EDF Regulation).
Eligible countries
All applicants are legal entities established in an EU Member State, Norway or Ukraine.
Ownership and control
No applicant may be controlled by a legal entity or authority from another country (Article 9 EDF Regulation); a Member State guarantee may allow an entity controlled from a non-associated country (Article 9(4)); those guarantees must address, among other things, independence in carrying out the action, protection of sensitive/classified information, ownership of results/IP and restrictions on third-country access. Applicants must accurately disclose the relevant information in their ownership and control declarations. Infrastructure, assets and executive management are located in an eligible country for the whole duration.
Legal personality
Natural persons are not eligible; applicants are incorporated (with a VAT number) before the deadline. Sole traders without separate legal personality may apply if they meet the other requirements.
One selected experiment per consortium and per organisation
A consortium may apply to several topics but only one of its experiments is selected; an organisation cannot take part in two selected proposals.
Funding ceiling per beneficiary
Up to EUR 60,000 per beneficiary, EUR 180,000 per experiment.
Proposals are submitted by the leading organisation, in a single stage, on the BattleVerse Evaluation Management System (EMS). Applicants register, create a new proposal (title, acronym, topic, keywords), write a summary of about 300 words in the Technical Proposal section, upload the proposal (PDF, 20 MB at most) and the mandatory Declaration of Honour and Ownership Control under Annexes, then press Submit. New versions can be submitted until the deadline; only the last one is evaluated. Late submissions, including force majeure, and proposals sent by any other means are rejected.
| Step | Date |
|---|---|
| Call opening (F&T portal) | 14 September 2026 |
| Submission deadline | 16 November 2026, 17:00 CET |
| Acknowledgement of receipt | By email after the call closure (not a proof of eligibility) |
| Eligibility check, evaluation, ranking | After the closure; every applicant receives the outcome by email |
| Sub-Grant Agreement with the coordinator (CERTH) | After selection; initial payment of 30% at signature |
| Experiment | 6 months (M1 to M6) |
| Criterion | Aspects | Threshold | Tie priority |
|---|---|---|---|
| Technical Excellence | Quality and soundness of the proposed solution; clear, ambitious and well-defined objectives (e.g. SMART); relevance to the sectors and technologies of the call; alignment with BattleVerse objectives and standards; level of innovation compared to current processes, products or services; robust and consistent methodology; consistent overview of the Technology Readiness Levels. | 3 | 2 |
| Impact | Clear technological and operational impacts; clear and realistic KPIs; capacity to enhance European strategic autonomy and competitiveness; capacity to contribute to the BattleVerse Innovation Hub; potential replicability of the experiment; exploitation model and exploitation potential. | 4 | 1 |
| Implementation | Overall work plan (deliverables, milestones, Gantt chart); capacity to carry out the experiment (infrastructure, technological expertise, previous experience); capacity and appropriateness of the team; coherent risk management strategy; alignment with European ethical and security standards (e.g. GDPR); budget, eligible costs and resource allocation. | 3 | 3 |
Each criterion is scored 0 to 5: 0 Fail, 1 Poor, 2 Fair, 3 Good, 4 Very Good, 5 Excellent. Only proposals above the threshold of each criterion and with a total of 11 or more are retained. Evaluators receive the proposal truncated to 15 pages.
Each participant may receive up to 2 individual one-on-one coaching sessions of one hour at most, on technical orientation and alignment with BattleVerse requirements. Beneficiaries remain fully responsible for developing, implementing and validating their solution. Section 2.3 of the proposal must describe the technical support required, aligned with the support available for the selected topic, and how the experiment will participate in and contribute to the BattleVerse Innovation Hub.
| Topic | Coaching available |
|---|---|
| Topic 1 | Trustworthiness of AI-generated recommendations (traceability, confidence indicators, hallucination mitigation, operator approval workflows); integration with the BattleVerse architecture (interoperability, API usage, data exchange formats); explainability and visualisation of AI outputs; KPI definition and evaluation methodologies. |
| Topic 2 | Synchronisation with BattleVerse Digital Twins and services; federated simulation integration; simulation interoperability (APIs, interfaces, data exchange); testing, verification and validation. |
| Topic 3 | Integration with BattleVerse simulation services (APIs, telemetry logs, scenario state exchange); human-in-the-loop integration; MLaaS/MSaaS training and validation workflows; evaluation methodologies, KPI assessment and operational validation. |
Beneficiaries engage with the BattleVerse Innovation Hub (Learning Expedition in Paris, Final Evaluation and Hackathon in Athens) and may join three online webinars on the defence market (go-to-market dynamics, funding opportunities, investor expectations in defence and dual-use markets). Contact with military advisers, planners or end users, reference scenarios and simulated assets cannot be guaranteed in advance.
Each experiment lasts 6 months (M1 to M6). Applicants define their own work packages in section 3.1 of the template (start and end month, lead, objectives, tasks, expected outcomes, deliverables), a Gantt chart and a list of milestones in section 3.2.
| Deliverable | Due | Content |
|---|---|---|
| D1 Experiment design and technical approach | M1 | Scope, methodology, system requirements, KPIs, data management strategy, timeline, expected activities and foreseen barriers. |
| D2 Technical implementation and demonstration | M4 | Implementation, integration and testing; architecture and technical specifications; interface or API specs, repository, deployment packages, integration or user guides; validation results and a live or recorded demonstration in the target environment. Source code is not required if KPIs and objectives are addressed, but a documented API service is expected for integration. |
| D3 Final experiment report and recommendations | M6 | Results, test scenarios and performance metrics, impact on the organisation and on the BattleVerse ecosystem, interaction with BattleVerse components, exploitation strategy, lessons learnt and recommendations. |
Selected experiments must demonstrate the capacity and resources to take part in these activities; the in-person sessions must be included in the budget.
BattleVerse is a research project in its first year: experiments integrate with a development environment that matures during the sub-grant, not a finished product, and no reusable framework is provided. Simulation components are federated with HLA (IEEE 1516, Pitch pRTI, object model based on RPR-FOM extended for BattleVerse); AI and decision-support components connect through a BattleVerse service API; DIS and FMI/FMU models integrate through a gateway or as HLA federates. No interface specification is available before submission: proposals describe their high-level integration pathway (HLA federate or API client). At onboarding (M1) selected experiments receive the integration guide, API specifications and SDK access; in the first two months they align with a BattleVerse use case, deliver an integration plan and complete a first connectivity test. Applicants provide their own environment for testing and validating the solution at TRL5.
30% after signature of the Sub-Grant Agreement; 70% after successful completion of the experiment and approval of all required deliverables and reports.
The call documents set no open-access or FAIR data obligation. Dissemination is expected during and after the experiment (social media posts, newsletters, technical publications, sector-specific events), and Topic 1 lists a synthesis of experimental results on benchmarks that can be disseminated within the community. Experiment data, even unclassified, may not be routed, processed or stored through cloud computing or AI services hosted outside the eligible countries.
The call is SME-led and technology-driven: a single SME may apply alone. Consortium partners are complementary capability providers (up to two), and non-SME partners such as RTOs or universities must be justified by the absence of a suitable SME. No multi-actor, transdisciplinary or SSH involvement is required; access to military end users through the consortium is not guaranteed, so applicants describe their own validation approach and may add letters of support from end users.
Selected applicants sign a Sub-Grant Agreement (SGA) with the project coordinator, CERTH, accepting the lump-sum financial support and its obligations (reporting, intellectual property, record keeping, payments). Its terms are predefined and cannot be negotiated; a draft will be provided. Beneficiaries notify the coordinator without undue delay of any change in ownership or control (acquisitions, mergers, changes of controlling interests); the consortium may reassess eligibility.
Progress is followed in individual monitoring meetings (M3 and M5) with the BattleVerse technical partners (progress against plan, technical and operational challenges, next steps), and the experiment is assessed at the Final Evaluation and Hackathon (M6) in Athens, where final results, lessons learnt and recommendations are presented. The final payment (70%) requires successful completion and approval of all deliverables.
TRL → 5
Grouped by technical topic — a proposal targets one branch.
Trustworthy, human-in-the-loop LLM decision support
LLMs and agentic AI for strategic decision support in multi-domain scenarios
On-premises LLMs for operational security and data sovereignty
Integration with MSaaS-compliant simulation environments
Symbolic reasoning, knowledge bases and retrieval-augmented approaches with LLMs
Game-theoretic modelling and AI-assisted scenario generation for wargaming
Planning under uncertainty
Order generation and explanation in natural language
Explainability, traceability and human approval of AI-supported decisions
Safeguards against hallucinated or unsafe recommendations
Surface and submarine warfare simulation
Multi-physics and multi-scale models of platform mobility
Modelling of physical and operational effects
Physical environment models
Interactive agents and behavioural models
Synchronisation and interoperability with Digital Twins and federated simulation services
Dynamic, adaptive and scalable simulation frameworks
Simulation assumptions on distributed situation awareness
Automatic planning for mission preparation, control and contingencies
AI-driven swarm coordination across heterogeneous UxVs
Distributed AI for communication, shared situation awareness and coordination
Human supervision and control of autonomous systems
ML pipelines for situation awareness and autonomous behaviours
MLaaS/MSaaS platforms for model development and validation
Operation in degraded, contested or GPS-denied environments
Operations against hostile agents in uncertain, constrained environments
Interoperability between UxVs, sensors, communication nodes and BattleVerse simulation services
Grouped by technical topic — a proposal targets one branch.
LLM and multi-agent AI prototypes for planning and decision support
Human-in-the-loop oversight of AI-assisted decision-making
Explainable, robust and safe AI outputs under uncertainty
LLMs integrated with symbolic reasoning, simulations, game theory and RAG
Performance evaluation of AI decision-support systems
Operational insights on LLM decision support in defence
Benchmark synthesis disseminated to the community
Prototype multi-domain simulation components or tools
Demonstrations of surface, underwater, electronic warfare or cross-domain simulation
Improved modelling of physical and environmental factors
Simulation agents or behavioural models
Executable simulation components with documented interfaces
Validated operational relevance and interoperability
Operational insights on multi-physics simulation
Prototypes for multi-UxV mission planning and coordination
Swarm coordination strategies based on ML or distributed AI
Validated human supervision of swarm operations
AI pipelines for training and evaluating coordinated behaviours
Resilient autonomy under degraded or contested communications
Operational insights on swarm-based autonomous systems
European strategic autonomy and competitiveness in defence
New SMEs and start-ups in the European defence sector
Enhanced mission effectiveness through European simulation add-ons
Regulation (EU) 2021/697 establishing the European Defence Fund
highEstablishes the European Defence Fund and its participation rules: eligible countries, ownership and control of participants (Article 9), definition of control (Article 2(6)) and security responsibilities.
Every participant is established in, and controlled from, the EU, Norway or Ukraine, with infrastructure and management located there; any non-EU shareholding is disclosed in the Ownership Control Declaration.
Commission Recommendation 2003/361/EC
highDefines micro, small and medium-sized enterprises by staff headcount, turnover and balance sheet, including linked and partner enterprises.
The lead applicant and any SME partner self-assess their SME status correctly; status may be checked by sampling.
Commission Implementing Decision C(2025) 568 final
mediumDecision of 29 January 2025 that, with the EDF Regulation, makes entities from Ukraine eligible alongside EU Member States and Norway.
Applicants from Ukraine are eligible on the same footing as those from the EU and Norway.
Regulation (EU) 2016/679 (GDPR)
mediumEU General Data Protection Regulation.
Section 3.5 explains how personal data, if any, is processed lawfully, minimised and protected; the Implementation criterion checks alignment with European ethical and security standards such as the GDPR.
European principles for human-centred AI
mediumTopic 1 requires solutions aligned with European principles for human-centred AI: explainability, robustness, transparency, accountability, safety and security, with humans retaining full control over decisions.
Topic 1 proposals show human-in-the-loop control, explainable and traceable outputs and safeguards against unsafe recommendations.
Regulation (EU, Euratom) 2024/2509 (EU Financial Regulation)
lowSets the exclusion situations and administrative sanctions that every applicant declares in the Declaration of Honour.
A signed Declaration of Honour stating that the organisation is in none of the exclusion situations.
Commission security rules for protecting EU classified information (Decision of 13 March 2015)
lowRules for handling EU classified information, applicable if an experiment involves it.
The proposal contains no classified information and states, with a brief justification, any need to access classified or security-restricted assets.
Talk to the Grant Coach to build your concept. There is no set order: start wherever your project starts. The sections below mirror what the conversation produces, and your coverage tracks the progress. You can refine everything once your project workspace is created.
The branch you're building for — it shapes what the call expects of you, and what your coverage is measured against.
The call requires you to pick or distribute several structuring elements before submission. These are locked at proposal creation time and drive the project backbone. Read-only here.
battleverse-oc1-topic
Topic 1: LLMs for strategic decision-making, symbolic reasoning, game theory, and mission planning under uncertainty
Topic 2: Multi-physics simulation capabilities for surface/submarine warfare, electronic warfare, and continuous combat across land, air, or naval platforms
Topic 3: Distributed AI for autonomous systems (UxVs) with for coordinated swarm operations
Deliverables (3) — imposed by the call
Experiment design and technical approach
M1Technical implementation and demonstration
M4Final experiment report and recommendations
M6Milestones (1) — imposed by the call
Experiment completed and all deliverables approved
M670%Fully managed by GrantForge
Click Create proposal above to start writing this form inside GrantForge. The structure below is what we generate for you.
Written in English on the official template and uploaded as a PDF. Sections 1 to 3 together must not exceed 15 pages, starting from section 1 (tables, figures and references included); excess pages are disregarded and evaluators only receive the truncated version. Hyperlinks meant to expand the proposal are ignored. Reference font Century Gothic, minimum 11 pt, single line spacing at least, A4, margins of at least 15 mm.
Before section 1: an identification table (title, acronym, leading organisation, type of entity, country of establishment, Partner I and II if applicable, technical topic) and a compliance self-assessment (Yes/No: Guide for Applicants accepted; GDPR compliance; no conflict of interest; not subject to control by a non-EU country, entity or nationals; consent to future contact).
| Section | Sub-sections | Indicative length |
|---|---|---|
| 1 Technical Excellence | 1.1 Experiment Overview; 1.2 Objectives of the Experiment (SMART); 1.3 Technological excellence; 1.4 Methodological Approach (with technical architecture); 1.5 Technological Maturity (start and final TRL) | 3 pages |
| 2 Impact | 2.1 Expected impact (with KPIs); 2.2 Relevance for the European Defence Sector (with target groups); 2.3 Collaboration with the BattleVerse Ecosystem; 2.4 Exploitation and Sustainability Strategy (with IP) | 5 pages |
| 3 Implementation | 3.1 Work plan (one table per work package); 3.2 Timeline and Gantt Chart (with milestones); 3.3 Personnel and Resource Management; 3.4 Risk Management Strategy (likelihood, severity, mitigation); 3.5 Ethics, Security and Data Protection; 3.6 Budget of the experiment | 7 pages |
Proposals are evaluated as submitted: there is no possibility for significant changes to content, budget or consortium during grant preparation.
Signed by the authorised representative of each organisation. It covers: accuracy of the application; legal status; compliance with the eligibility criteria for the whole activity; the ownership and control declaration (legal name, PIC if any, legal registration number, place of establishment, SME or other, owners holding 5% or more of capital or voting rights or specific rights, with ID copies); either a declaration of control by a non-associated third country (with the exception of Ukraine) or a declaration of no control, including that infrastructure, facilities, assets, resources and executive management are located in a Member State, an associated country or Ukraine for the whole duration; involvement in other EDF or EU-funded projects; commitments (resources, funding, ethics and research integrity, EU classified information rules) and the exclusion situations of the EU Financial Regulation. A national company registration number is accepted where no VAT number is requested; applicants must nevertheless be incorporated and have a VAT number before the deadline.
No separate budget form: section 3.6 of the proposal holds a simplified table with, for the leader and each partner, personnel costs, other costs (materials, equipment, travel), total cost and total funding requested, plus the total funding requested for the experiment. Figures must be consistent with the work plan and include only costs exclusively dedicated to the experiment.
Canonical EU Funding & Tenders Portal page for this call.
https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/competitive-calls-cs/15861
Institutional / marketing website of the parent programme.
https://battleverse.ems-carsa.com/call/detail/1
Where you submit your proposal (F6S, opencalls.fund, EPSS, MS Forms…).
https://battleverse.ems-carsa.com/call/detail/1