Belgium’s Defence Innovation Funding: A Map for Companies

by | Nov 8, 2026 | Defence & Security, Commentary, Publications

Belgium's Defence Innovation Funding A Map for Companies

A Belgian company looking for defence innovation funding can work with four federal channels, alongside regional instruments in Flanders and Wallonia and a set of sector platforms. They do different things, and a company may engage with more than one, subject to each instrument's rules.

Three of the four federal channels took their current shape recently. Start2DEF launched in October 2025. ODIN was launched in January 2026 and sits inside the Defence, Industry and Research Strategy (DIRS), which the Royal Higher Institute for Defence (RHID) presented in 2022. SFPIM Defence was created by royal decree in July 2025, and its strategy and funding were approved in April 2026. DEFRA is the oldest of the four and has run an annual call since 2021, managed with the Belgian Science Policy Office (BELSPO).

Belgian companies had other options before this, including regional innovation support and the EU programmes, and Flanders and Wallonia have since built defence funding of their own. What has changed is the amount of money carrying a specific defence mandate, and the number of separate bodies deciding where it goes. The channels differ in funding mechanism, decision process and time horizon, and their budget relationships are not fully mapped here.

This post maps the four federal channels, the regional instruments in Flanders and Wallonia, and the sector platforms that work with federal Defence. It says what each one funds and sets out what is still unclear. It is written for companies and research teams deciding where their work fits.

Four federal channels compared

Route

Who runs it

Form of support

Published amount and basis

DEFRA

RHID manages and funds; BELSPO manages the calls

Defence-related research projects, chosen by theme

€26 million indicative for the sixth call; the four 2026 call envelopes total €51 million

ODIN

Ministry of Defence

Innovation and industrial partnerships; access depends on the relevant mechanism

€3.7 billion announced for 2026 to 2035

SFPIM Defence

SFPIM, the federal holding company

Equity stakes in defence and dual-use companies

Conditional drawing rights, potential Recovery and Resilience Facility resources and transferred holdings; see text

Start2DEF

RHID

Challenge-linked voucher for proof-of-concept work by startups and small firms

€20,000 per selected company

These figures are not directly comparable and should not be added together. The table sets a call budget beside a ten-year announcement, an investment capacity and a per-company voucher.

The differences matter more than the totals. DEFRA pays for research against themes it sets in advance. ODIN commits money over several years to partners it selects. SFPIM Defence buys shares in companies. Start2DEF looks for early ideas against named problems.

So the right question is not which one is biggest. It is which one is buying the kind of thing your organisation produces.

DEFRA: the research call

DEFRA is Belgium's national defence research programme. BELSPO organises the calls and RHID funds the selected projects. For the sixth DEFRA call in 2026, a proposal had to come from a network with at least one public or private non-profit research institute and at least one private company. Companies needed legal personality, a registered office in Belgium and operational or research activities on Belgian territory. Foreign organisations could join only as non-funded partners. Check the next call's rules before assuming the same terms apply.

The money does not cover everything. The 2026 information document limits funding for private companies to 65 per cent of eligible costs, rising to a maximum of 80 per cent depending on company size, while public research institutes and private non-profit research centres can be funded up to 100 per cent. A company therefore needs to plan its own contribution; alongside the documentation and security verification the call requires.

The 2026 cycle was not a single call. Four rans in it: the sixth DEFRA call with an indicative €26 million over 10 themes, a Space Call with €16 million, a Human Factors Open Call with €4 million, and an open call on Pulsed Directed Energy Effectors with €5 million. Those four envelopes total €51 million, which is a sum of published call budgets rather than a named cycle budget or a record of what was awarded. Separate thematic calls can be much larger than the yearly ones. The NGCAT call of 2024, a thematic call of its own, was worth €60 million over 4 themes and funded 15 projects.

According to NCP Flanders' notice of 3 November 2026, the launch of the 2027 DEFRA calls has been postponed to spring 2027, and the information days planned for 23 and 24 November 2026 are postponed with it. No themes or budgets for 2027 had been published when we checked on 8 November 2026. We wrote about that postponement in a separate commentary.

The part of DEFRA that gets least attention is the call for themes. Ahead of the 2026 round, RHID invited Belgian universities, research institutions and companies to propose research themes, and used those submissions when it set the themes that were funded. The submission deadline was 10 June 2025. The sixth call that followed contained nine specified themes and one open category. A theme submission neither guarantees inclusion nor secures a later award.

The selection criteria are published. RHID weighs its own research areas, the technology domains in the DIRS framework, the multi-year outlook of the European Defence Fund, and the priorities of Belgian Defence staff. In other words, a company that wants DEFRA to fund its field can argue for that field a year in advance, and the argument lands better if it is framed in those terms.

ODIN: the innovation programme

ODIN is the organisation for defence innovation and the defence industry. Belgium launched it in January 2026, and when Defence Minister Theo Francken set out the detail in June, Belga reported an announced €3.7 billion programme for 2026 to 2035. Earlier reporting put the figure at €3.6 billion. That is an announcement covering ten years, not money already committed or paid.

ODIN sits inside DIRS, the framework RHID presented in 2022. Belga reported that under that strategy about €350 million a year goes to research, innovation and industrial production. That is a separately reported DIRS figure, not money additional to the ODIN announcement.

The clearest example of how ODIN works is the imec agreement. imec and Defence announced a five-year strategic partnership on 30 October 2026. The Brussels Times, reporting with Belga, put ODIN support at €10 million a year, about €50 million in total, for work on secure communications and on compact radar and laser systems against drones. That total is an announced commitment, not money already paid.

The imec agreement shows that ODIN can support multi-year partnerships. It does not establish that all ODIN support works that way. A company should ask Defence or RHID which published calls, partnership arrangements or supplier opportunities apply to its own capability.

SFPIM Defence: the investment fund

This route is not a grant programme at all. SFPIM Defence is a specialised investment subsidiary of SFPIM, the federal holding and investment company. Its creation and statutes were approved by royal decree of 28 July 2025. On 3 April 2026 the Council of Ministers approved its investment strategy, its funding, the participations transferred into it, and its governance and ethical framework, and the release was published on 4 April. Approval of its creation is not the same as operational readiness.

The fund works along two lines: anchoring key players in the sector, and investing in innovation. These are organised around 14 key investment themes. The April Council release refers to them but does not list them, so confirm the current set with SFPIM Defence.

Its funding parameters were set out in parliament. In the Chamber's Finance and Budget Committee on 15 September 2026, Jan Jambon described an annual drawing right of up to €250 million from SFPIM over the next five years, an additional drawing right of up to €300 million payable if a major asset is realised during the five financial years ending 31 December 2030, and up to €150.8 million in unallocated Recovery and Resilience Facility funds during the fund's first year. SFPIM also transfers its defence, aerospace and cybersecurity holdings into the subsidiary. These are financing parameters, not proof that a given total is available now. The €2 billion figure that circulates in press reporting appears in the question put to the minister, not in his answer.

Its first holdings come from the parent company. The Council of Ministers ended SFPIM's delegated mandates for BeLightning and John Cockerill Defence so that those participations could be contributed to the new fund.

SFPIM is the federal holding and investment company and describes itself as Belgium's sovereign wealth fund. Its corporate site reported stakes in 189 companies and about €14 billion under management on behalf of the federal state when we checked on 8 November 2026. That parent portfolio is not the amount available through SFPIM Defence.

For a company, this is a different conversation from a grant. The fund takes a stake, so an approach needs a commercial and strategic investment case alongside the technical one. Clarify the proposed financing instrument, governance terms, time horizon and return expectations with the fund directly. A good research idea is not by itself an investment case.

Start2DEF: the open door for startups

Start2DEF is a small-scale challenge programme for Belgian startups and small and medium-sized enterprises (SMEs). RHID runs it, with Deloitte delivering the programme, and it asks for proof-of-concept ideas against named defence problems.

Applications for the current edition opened on 4 November 2026. An online information session is scheduled for 12 November at 16:00, and sign-up for that session closes on 11 November at 20:00. The expression-of-interest deadline is Thursday 26 November 2026 at 20:00 CET, as published on the programme details and submission pages. Registering for the information session is not the same as submitting an application.

The submission page asks for a completed expression-of-interest form on the obligatory template, carrying a proof-of-concept proposal, together with a three-minute recorded video, ultimate beneficial owner (UBO) records and the company number.

The current edition sets two challenges:

  • Resilient and modular communications for unmanned systems
  • Adapted nutrition for wounded soldiers during medical evacuation

The currently posted programme FAQ states the following. It still links presentations from the October 2025 information session, so confirm these terms against the 2026 information document before committing:

  • The programme is only for Belgian companies with their headquarters in Belgium, and a Belgian subsidiary or brand of a foreign company does not qualify. The FAQ does not address foreign ownership of a Belgian-headquartered company. Sole traders with a Belgian enterprise number can apply.
  • No consortia. Applicants enter as a single company.
  • A company may submit more than one application but can only be selected once.
  • Selected companies receive a voucher of €20,000. Anything above that is at the company's own cost, and random audits are possible after the challenges close.
  • The challenge runs for eight weeks at roughly one day a week, with bootcamps expected in weeks 2 and 5.
  • Deloitte and RHID review applications together with end users. RHID takes the final decision. The working language is English.
  • The FAQ does not yet set out IP ownership. It points to an information document marked as coming soon. Clarify ownership, usage rights and payment conditions before committing resources.

One point to watch. Some material still online describes the challenges of the previous edition, which covered drone swarms and cybersecurity for critical infrastructure. Work from the current challenge statements, not from last year's slides.

For a small company with no defence track record, this is a possible entry point. The voucher is modest, and the value is in feedback from defence end users and in the experience of developing a proposal against an operational need. Participation does not itself establish an RHID endorsement or a procurement commitment.

Regional instruments and sector platforms

The federal channels are not the whole picture, and two different things sit alongside them. Flanders and Wallonia run their own regional funding instruments, with their own budgets and their own eligibility rules. Separately, sector platforms coordinate domain work with federal Defence and RHID. Where a platform's coordinator is based does not make the money behind it regional and does not by itself decide who may take part.

Flanders funds through VLAIO. The Flemish Defence Plan was approved in April 2025, and VLAIO runs a continuous call for defence projects under it. The 2026 defence call accepts applications on a rolling basis. It funds development and research projects, and purely military as well as dual-use applications qualify, provided the military application is a significant part of the business case. Base support is 25 per cent for development projects and 25 to 50 per cent for research projects, depending on the share of research. Small companies receive a further 20 percentage points and medium-sized companies a further 10. A further 15 percentage points requires qualifying balanced collaboration and is not automatic. The caps are 60 per cent per company partner for development and 80 per cent for research. Classified material is handled separately, not through the online portal, and VLAIO offers an optional pre-discussion before you apply. Confirm current availability and the project-start rules before incurring costs.

That money is already moving. In March 2026 the Flemish government awarded €3.65 million to eight companies, covering autonomous navigation, drone and intruder detection, composite aircraft parts, radar calibration, underwater mine-detection robots, 3D-printed metal parts, training games and AI that detects online extremism.

Wallonia funds through SPW Recherche. COOTECH Défense launched in March 2026. It is open to companies only, for collaborative project portfolios whose economic impact has been assessed in advance. Projects must fit Defence priorities, specifically the DIRS framework, or develop dual-use products. Projects must start at technology readiness level (TRL) 3 or above and target at least TRL 6 by completion, on the 1 to 9 scale. The launch notice does not publish rates or deadlines, so ask SPW Recherche directly.

Brussels is a separate jurisdiction. Innoviris states that no new projects can be funded in 2026, that no new calls will be launched and that all programmes are suspended for the year. Projects already under way continue under their signed decisions and agreements, and applications already submitted will still be processed and may be eligible for funding in 2027 depending on available resources. A Brussels-based company should check eligibility and current availability with Innoviris directly.

The sector platforms are a different kind of thing. They coordinate domain work with federal Defence and RHID rather than distributing a regional budget, so read them as routes to partnership and to early information and check the funding instrument sitting behind any particular call.

Land. CAP4LAND is a strategic programme developed in partnership with Belgian Defence and coordinated by Pôle MecaTech, the Walloon cluster. It builds a technology roadmap for the land domain and runs thematic working groups where priority projects are identified. MecaTech's wider defence and security portfolio includes the M4 multigauge machine gun coordinated by FN Herstal, the IRIS surveillance project and the ElectroHOB rocket concept. The 2026 Try-Out Call had two tracks: CAP Project, for collaborative innovation at TRL 6 to 7, and CAPImpact Défense, for operational excellence in Belgian manufacturing SMEs. Their letter-of-intent deadlines were 14 June and 5 July 2026 and have passed. CAPImpact offers support worth €75,000 of which the programme co-finances 70 per cent, delivered as expertise and services rather than as a cash grant. Ask MecaTech about later intakes.

Sea. In October 2025 Blue Cluster published a feasibility study for a maritime security ecosystem, commissioned by the Ministry of Defence and RHID under DIRS. That work led to Blue Bastion, run by Blue Cluster with RHID. It has two tracks: demonstrator projects of up to €350,000 for up to six months, taking a technology from TRL 5 or 6 to TRL 7, and research projects of up to €2 million for up to 24 months. Both target maritime security against cyber, hybrid and physical threats, and dual-use applications are explicitly of interest. No Blue Bastion call is open at the time of writing. The second call closed on 1 October 2026, and Blue Cluster says the third will launch in spring 2027. The full-proposal deadline of 1 December 2026 applies to research applicants already selected at the pre-proposal stage, so it is not a fresh entry point.

Air and space. We found no single platform for the air domain equivalent to CAP4LAND. Skywin, the Walloon aerospace cluster, is a relevant contact for Walloon aerospace and space companies and publishes Space4Defence material. Flanders has its own Space4Defence orientation roadmap under the Flemish innovation and industry strategy for security and defence. Use the network that matches your location and technology and check the specific funding instrument behind it.

One organisation is worth separating out. Flanders Make has a defence and security track covering autonomous drones and unmanned vehicles, fleet life-cycle management, teaming technology and distributed microfactories, and it has a memorandum of understanding with NATO's DIANA programme, which gives selected DIANA innovators access to its test facilities. We found no defence call of its own on its site. It can provide research and testing capability, and its membership offer describes access to its research team and priority access to the results of its research projects. What that means in practice depends on the collaboration, service or programme terms, so treat it as a capability partner rather than a funding route.

Two things follow. Regional funding instruments are anchored in Flanders, Wallonia or Brussels, so where your company sits decides which authority applies. The sector platforms are not bound the same way, and a platform's location does not by itself decide who may take part. Beyond that, these instruments differ in eligible activities, award size and timetable. Compare the current call conditions rather than assuming that regional or cluster-managed support is smaller or faster than the federal routes.

Where the EU layer connects

None of these sit apart from the European programmes, though who receives the money differs from one to the next. The European Defence Fund (EDF) supports collaborative defence research and development. EUDIS is an innovation scheme within EDF, with measures aimed at smaller firms and new entrants. The European Defence Industry Programme (EDIP) covers industrial reinforcement and common procurement support. SAFE lends to member states, so a company reaches the related work through procurement and supply chains rather than by applying for a SAFE loan itself.

The connection is not only thematic. RHID says it weighs the multi-year outlook of the European Defence Fund when it selects DEFRA themes. EDF priorities inform those themes rather than determine them, but a company that follows EDF work programmes still gets a useful early read on where DEFRA may go.

RHID states the link directly. It presents DEFRA projects as leverage for research institutions and industry to improve their technological level and increase their chances of participating successfully in other programmes, naming the European Defence Fund. Our own observation, from working inside EDF consortia, is narrower and sits alongside that: national delivery experience is also something consortium coordinators weigh when they choose partners. That second part is a judgment from experience, not a published rule.

There is a caution attached. Eligibility is instrument-specific. EDF generally requires at least three independent eligible entities from three different member states or associated countries, while calls for disruptive technologies can work with smaller consortia. Other instruments and activities set their own applicant rules. The administrative load on a collaborative European application is heavier than on a national call, and earlier national delivery can strengthen a proposal without being a universal prerequisite.

What is still unclear

Four things are not settled, and anyone planning around this map should know which parts are firm.

The division of labour between DEFRA and ODIN. Both fund defence research and innovation. We checked the BELSPO and RHID DEFRA pages and the DIRS programme's own site. DEFRA is presented there as part of the DIRS toolbox, and ODIN is not mentioned on that page at all. The pages we cite here do not give a consolidated account of how the two divide responsibilities, or of whether DEFRA will be reshaped around ODIN.

The 2027 DEFRA timetable. The calls are postponed to spring 2027 and no themes, budgets or dates have been published. For comparison, the sixth 2026 call ran from a pre-proposal deadline of 19 February 2026 to projects starting in December 2026. A spring 2027 launch could shift the later stages, but the revised calendar is needed to know whether project starts move or the assessment period is shortened.

The next call for themes. The 2026 themes came from submissions that closed on 10 June 2025. We looked on the RHID DEFRA pages, on BELSPO and on the DIRS site, and found no theme invitation for the 2027 or 2028 rounds in those sources. That is not proof that no consultation is planned. Confirm the next one with RHID directly.

The budget setting. The government's Strategic Level Report, approved by the Council of Ministers on 6 February 2026, maintains defence expenditure at 2 per cent of GDP from 2025 through 2033 and moves to 2.5 per cent from 2034. Separately, NATO Allies committed at The Hague to 5 per cent of GDP by 2035, made up of at least 3.5 per cent for core defence requirements and up to 1.5 per cent for broader defence and security investment. These are plans and commitments, not spending already made. Against that, the federal government is negotiating a fiscal adjustment reported at around €10 billion by 2029, combining spending cuts and revenue measures. The Prime Minister offered his resignation on 2 November 2026. The King rejected it on 6 November and asked him to resume budget negotiations and report back on 16 November.

The unresolved federal budget creates uncertainty for future authorisations, but the effect must be checked instrument by instrument, because they draw on different budgets and different decision processes. It is worth separating announced envelopes, authorised commitments, signed awards and payments. Treat a delay, a cut or an unchanged allocation as established only when the responsible authority says so.

Reading the map for your own organisation

These routes buy different things, so the first question is what you are actually offering.

If you are

Possible route to investigate

What it asks of you

A startup or small SME with an early idea

Start2DEF

A Belgian HQ, one company, challenge fit and the required expression-of-interest package

A research team or a company with a research partner

DEFRA

A project against a published theme, with at least one research institute and one company

An organisation with a relevant technology or industrial capability

ODIN

A capability Defence wants to commit to; ask which mechanism applies

A company that needs capital, not a grant

SFPIM Defence

An investment case, not only a technical one

A company in the land, maritime or aerospace domain

A regional funding authority, or a sector platform

For regional funding, a base in that region; for a platform, a fit with its roadmap

This table points to a route worth investigating rather than settling eligibility. Sector, location, maturity, ownership and whether a call is currently open can all change the answer, so check the current conditions of the instrument itself. Intake dates here were checked on 8 November 2026.

Three things are worth preparing whichever route you take.

First, know which end user your work is for. Identify the relevant military or civilian end user and show how the proposed work fits the objectives of the instrument you are approaching.

Second, be honest about where your capability ends. Naming the part you cannot deliver, and the partner type who can, is what makes the rest credible. Consortia and investors both read overreach quickly.

Third, for thematic research calls, watch the upstream moment and not only the deadline. By the time such a call is published, its themes were set months earlier. For rolling support and for investment approaches, confirm the applicable process directly with the body concerned.

What we will cover next

This map is a starting point, and several parts of it deserve their own treatment. We will keep writing on them.

ODIN and DEFRA. The largest announcement and the oldest call now sit side by side with no published division of labour between them. That question decides a lot about where Belgian defence research money goes after 2027.

The call for themes. The 2026 DEFRA cycle shows why early theme consultation matters. It deserves a piece of its own, and we will look at what happens to the 2027 and 2028 cycles.

The regional instruments. Flanders and Wallonia both built their own defence funding during 2025 and 2026, with different rules and different entry points. That shift happened while the federal picture was changing, and the two are worth reading together.

Before committing resources, confirm eligibility, the current intake dates, your own contribution, the IP terms and the route from research support to procurement. These programmes move faster than their websites. If you are working through one of them and something here is already out of date, we would rather hear it than repeat it.

Sources

All pages were consulted on 8 November 2026.