Commercialize Deep Tech in Poland: NCBR STEP Grants for Hardware, Quantum, Robotics, Cleantech and Biotech Ventures

Commercialize Deep Tech in Poland: NCBR STEP Grants for Hardware, Quantum, Robotics, Cleantech and Biotech Ventures

If you build deep-tech hardware, software, or biotech and you are trying to move from proof-of-concept to real-world deployment, Poland’s STEP instrument is one of the largest sources of non-dilutive money in the region. STEP sits inside the FENG programme (Fundusze Europejskie dla Nowoczesnej Gospodarki) and is run by NCBR, the National Centre for Research and Development.

STEP is not a single competition with one annual deadline. NCBR splits it into three sectors — digital technologies and deep-tech, clean and resource-efficient technologies, and biotechnology — and each sector runs two separate calls: Track A for critical technologies with commercial potential on the EU internal market, and Track B for projects that reduce or counter the EU’s strategic dependencies. That means six STEP windows in 2026, staggered across the year, plus the related Ścieżka SMART calls for consortia.

The practical consequence: missing one window does not mean waiting a year. It usually means waiting a few weeks for the next sector to open.

Where the 2026 cycle stands right now

NCBR publishes a consolidated 2026 competition schedule, which is the single most useful page to bookmark. As of late July 2026:

Two earlier Ścieżka SMART rounds have already run this year — one for individual large firms (23 Mar – 22 May 2026) and one for consortia (9 Apr – 12 Jun 2026), each with a 350,000,000 PLN budget. NCBR also ran a STEP IPCEI call from 30 Jan to 25 Feb 2026 with a 120,000,000 PLN allocation.

If your project is digital or deep-tech in the narrow sense, both 2026 windows have closed and you are now planning for the 2027 schedule. If your work touches clean technologies or biotechnology, you have an open call in front of you today.

The money involved

The figures published on each 2026 STEP call page are consistent across sectors:

• 150,000,000 PLN allocated per call.
• Up to 140,000,000 PLN of requested funding for an entire project — this is a ceiling, not a target.
• Grants (dotacja) as state aid for enterprises; research organisations are supported outside the state-aid framework for their non-economic activity.

STEP Ścieżka A documentation has also carried a floor on project size — around 7,000,000 PLN of eligible costs, rising to roughly 10,000,000 PLN where a large enterprise participates in a consortium. Those thresholds date from the 2025 edition and are reported for the 2026 Ścieżka A calls, but they are not restated in the summary text of every 2026 call page. Read the Regulamin wyboru projektów for the specific call you are targeting before you size your budget around them.

The parallel Ścieżka SMART consortium round is a better fit for smaller projects: a 3,000,000 PLN minimum in eligible expenditure, a 350,000,000 PLN pot, and the same 140,000,000 PLN per-project cap.

This matters because sizing is the most common reason a good project is filed into the wrong competition. STEP is designed for expensive, multi-year industrial R&D. If your realistic scope is two million złoty, Ścieżka SMART or a regional programme will treat you better.

What the money is actually for

Two things are easy to miss when reading these calls: scale and scrutiny.

Scale. The support finances high-cost, high-complexity work that ordinary startup spending cannot absorb — prototype hardening, custom fixtures and equipment, external testing, specialised hardware and software development, and validation in operational environments. If your product needs rigorous demonstration before customers will pay, this is the phase STEP was built to fund.

Scrutiny. Evaluation is two-stage. Stage I is a document review by an expert panel, with a chance to correct or supplement the application. Stage II includes a meeting with the applicant. NCBR allows up to 150 days from the close of a call for the whole process, with roughly 80 days for Stage I. Teams that are conceptually strong but operationally vague rarely survive the Stage II conversation.

Note also that implementation of results happens outside the project, funded by the beneficiary or other sources. The grant pays for the R&D that makes deployment possible; it does not pay for the deployment itself.

Who can apply

The applicant rules are the same across the 2026 STEP calls:

• A standalone enterprise, or a consortium of at most three members.
• Consortium members may be SMEs, large enterprises (including small mid-caps and mid-caps), research organisations, or non-governmental organisations.
• The consortium lead must be an enterprise. Research organisations and NGOs can only be partners, never leads.

Track B calls are framed around consortium delivery and strategic-dependency reduction, so read the specific call before assuming a solo application is accepted there.

Profiles that usually fit

A hardware control platform for precision manufacturing that needs industrial partner pilots and environmental qualification studies.

A robotics team with a fleet prototype that must be upgraded for industrial-grade reliability testing and field deployment.

A quantum-signal processing venture with initial demonstrators that needs pilot-grade hardware and named early adopters.

A cleantech company with a working process that needs scale-up trials, emissions measurement, and third-party verification before an industrial customer will commit.

A biotech spin-off moving from lab protocol to a pre-commercial productized method with planned validation and compliance steps.

Profiles that usually do not fit

A concept-stage service idea with no technical depth. A team with no test data, no validated prototype, and no defensible budget logic. A team that cannot prove ownership or licensing rights to core algorithms, sensor designs, or protected results. A project with no concrete market destination.

What “deep-tech” means in the digital sector

For the digital and deep-tech sector calls, NCBR’s topic attachment divides eligible themes into sectors and technology families, and expects a genuine deep-tech project to combine at least two critical technologies from different STEP areas. The published list spans advanced semiconductors; AI and high-performance compute; quantum computing, communication, and metrology; cybersecurity and digital identity; advanced connectivity and navigation; advanced detection; robotics and autonomous systems; and biotech-adjacent innovations.

Because that list is broad, belonging to a named sector is not the filter. The filter is specific technical novelty plus an executable route to market. Proposals that stay generic score poorly even when the topic line looks impressive.

Eligibility gates, in order

Gate 1 — formal fit. Standalone or consortium of up to three, enterprise lead. Many weak applications fail purely because the consortium structure is inconsistent with the rules.

Gate 2 — scope fit. The project must sit in the sector the call covers, and must not be a purely military solution. If your output is dual-use, justify the civil share and the commercialization model explicitly; reviewers check this.

Gate 3 — finance fit. Requested funding must stay under 140,000,000 PLN, and the project must clear the minimum eligible-cost threshold for the call. If your project is materially below the floor, Ścieżka SMART is the better route.

Gate 4 — legal context. Ownership and control arrangements must satisfy programme rules on beneficiary status and implementation territory. Get legal sign-off before drafting technical sections, not after.

Gate 5 — commercialization confidence. A project that produces only research output, with no route to adoption and no pilot demand, is too weak for STEP regardless of technical merit.

A readiness scorecard

Before committing weeks to a submission, answer these honestly:

• Can we write a technical plan with clear acceptance criteria?
• Do we have at least one realistic pilot or LOI-worthy target?
• Can we identify and cost at least six months of hard technical work?
• Are our core IP rights clear and documented?
• Do we have the internal execution capacity, or credible partners, to deliver the milestones?
• Do we have a finance lead who can map cash flow and defend the assumptions?

Mostly “yes” means you are a plausible applicant. Mostly “not yet” is a readiness problem, not a programme mismatch — and the fix list is the same either way.

Application process

1) Pull the exact call documents. Use the current call page on gov.pl for your sector and track. Do not work from stale PDFs or third-party summaries; NCBR updates documentation mid-call (the digital Track B call received a documentation update after announcement).

2) Write the problem statement first. One paragraph, plain language: what business pain exists, who pays, why existing solutions fall short. Then tie each technical feature to a measurable outcome. Without this, the proposal reads like a research paper.

3) Build the project map. Technical development steps, validation tests, commercialization milestones, and dependencies — vendors, labs, pilot customers, certifications. Fewer high-confidence milestones beat many vague ones.

4) Draft the budget before the narrative. Qualified cost structure, vendor quotes, and a realistic staffing plan decide credibility more often than prose does. Include contingencies you can justify.

5) Assemble the legal package early. Company registration documents, IP ownership proof or licence agreements, SME status declarations where applicable, and partner letters with defined roles for consortium applications.

6) Write for a reviewer, not for a pitch deck. Define acronyms on first use. Keep the focus on what changes because of the project, and why it would not happen without the support.

7) Run a pre-submission review. One technical reviewer, one commercial reviewer, both instructed to attack the assumptions. If neither can find two measurable KPIs and a reasoned commercialization funnel, rework.

8) Submit through LSI before the cut-off. Applications go only through NCBR’s system at lsi2.ncbr.gov.pl. Most 2026 calls close at 16:00 on the stated day. Submit early and keep exported files and screenshots.

Planning calendar

Working backwards from a close date:

• D-120 to D-90: confirm topic fit, project scope, consortium setup.
• D-90 to D-60: finalise technical architecture and budget assumptions.
• D-60 to D-30: prepare the legal and IP package, draft milestones.
• D-30 to D-14: complete the draft, request two external reviews.
• D-14 to D-7: tighten language, unify metrics, verify documents.
• D-7 onward: test portal login and signature flow, submit early.

For Biotech Track A closing 7 September 2026, that puts serious preparation start in May. If you are reading this in late July with nothing drafted, Biotech Track B (16 September) or the Ścieżka SMART consortium round (16 October) are the realistic targets.

Required materials

The binding list lives in the call templates, but the recurring core is:

• application form and project description,
• technical concept and development method,
• milestone and work-package plan,
• budget and financing model,
• multi-year financial and operational forecast where required,
• IP documentation and chain of title,
• team members and roles,
• SME status declarations and required legal confirmations,
• letters documenting partner or customer intent,
• supporting evidence from pilots, prototypes, or tests.

Treat appendices as an evidence vault: raw lab results, vendor offers, partner statements. Keep the main body readable and decision-focused.

What evaluators reward

Technical realism with measurable targets. A commercial pathway with early demand signals. Coherent budget logic. A team mix that covers both engineering and execution. Legal clarity on IP and ownership. They do not reward hype — the proposal has to say who does what, by when, and with what proof.

Common mistakes

Overstating readiness. Claiming advanced maturity while supplying only concept language creates an immediate trust gap. Anchor every claim to data, even partial data.

Too many workstreams. Poor sequencing reads as execution risk. Reduce scope and show dependencies.

No commercialization evidence. Add pilot intent, LOIs, and acceptance criteria that mirror buyer reality.

Unclear consortium roles. Inconsistency between lead and partner responsibilities causes administrative problems. Produce a governance and responsibility map before drafting.

Unvalidated budget. Numbers without quotes or stated assumptions read as guesswork.

Ignoring dual-use scrutiny. If military application is possible, explain the civil-use share and the commercialization logic explicitly.

Underestimating reporting. This is long-cycle funding with milestones and reporting obligations. Assign compliance ownership at the start, not at the first report.

Frequently asked questions

Which STEP calls can I still apply to in 2026?

Cleantech Track A closes 31 July 2026 at 16:00 and Track B closes 7 August 2026 at 16:00. Biotech Track A closes 7 September 2026 at 16:00 and Track B closes 16 September 2026. The Ścieżka SMART consortium round opens 7 August and closes 16 October 2026 at 16:00.

What happened to the digital and deep-tech calls?

Both 2026 windows have closed — Track A ran 14 April to 17 June 2026 and Track B (FENG.05.01-IP.01-003/26) ran 21 April to 26 June 2026. Applications are under evaluation. Watch the NCBR schedule page for the 2027 dates.

Can startups apply alone?

Yes, in the Ścieżka A calls, a standalone enterprise can apply. A consortium is allowed with up to three members, and the lead must be an enterprise. Track B calls are structured around consortium delivery — check the specific call.

How much can one project receive?

Requested funding cannot exceed 140,000,000 PLN for the whole project, against a 150,000,000 PLN budget per STEP call. There is also a minimum project size in Ścieżka A; confirm the current figure in the call’s Regulamin.

How long does evaluation take?

Up to 150 days from the close of applications, with roughly 80 days for Stage I. Stage II includes a meeting with the applicant.

Can foreign collaborators be involved?

Foreign collaboration is possible, but implementation and ownership must satisfy the eligibility and control rules. Get legal review before filing if this is uncertain.

What if the evaluation goes against us?

The call rules include a formal protest procedure for negative outcomes. Check the protest chapter and its timing rules in the current Regulamin.

Official links

• NCBR 2026 competition schedule: https://www.gov.pl/web/ncbr/harmonogram-konkursow-2026
• STEP Cleantech, Track A (004/26): https://www.gov.pl/web/ncbr/czyste-i-zasobooszczedne-technologie-feng-0501-ip01-00426
• STEP Cleantech, Track B (005/26): https://www.gov.pl/web/ncbr/czyste-i-zasobooszczedne-technologie-feng-0501-ip01-00526
• STEP Biotech, Track A (006/26): https://www.gov.pl/web/ncbr/step-sciezka-a-biotechnologie-nabor-feng-0501-ip01-00626
• STEP Biotech, Track B (007/26): https://www.gov.pl/web/ncbr/step-sciezka-b-biotechnologie-nabor-feng-0501-ip01-00726
• Ścieżka SMART, consortia (003/26): https://www.gov.pl/web/ncbr/feng0101-ip01-00326-sciezka-smart-projekty-realizowane-w-konsorcjach
• STEP digital and deep-tech Track A reference page (2025 edition, concluded): https://www.gov.pl/web/ncbr/step-sciezka-a-technologie-cyfrowe
• SME qualification tool: https://kwalifikator.parp.gov.pl/
• LSI submission system: https://lsi2.ncbr.gov.pl/

What to do this week

Pick one product line and one measurable commercialization target. Map your current technology maturity against the milestones you would promise. Collect proof artefacts for every technical claim you intend to make. Draft the budget categories and get at least one vendor quote per major cost line. Confirm your IP ownership documents. Then write a first pass of the evaluation narrative in plain, review-friendly language.

The process rewards method over inspiration. Teams that treat the application as a forcing function for real project planning tend to both score better and execute better afterwards — which is the part that matters once the money arrives.