By the Airvolt Team
Introduction: The Paradox of European Innovation
Europe says it wants innovation.
It invests billions through Horizon Europe and the European Innovation Council Accelerator (EIC) to fund “high-risk, high-impact” technologies.
In 2023, Airvolt’s AMPS (Air Monitoring & Purification System) was accepted into this framework — a project merging air purification, IoT air-quality sensing, and digital advertising into a self-financing clean-air infrastructure.
We were evaluated twice by the EIC in early 2024.
Both evaluations recognized the same product, the same team, and the same objectives.
One awarded 6 out of 12 points and called AMPS “groundbreaking.”
Two months later, the next panel awarded 0.00 points — even though its own evaluators wrote that our plan was “thorough, logical, and realistic.”
A zero, despite pages of positive feedback.
That isn’t a judgment — it’s a violation of procedure.
This article documents exactly how that happened and why it matters — not just for us, but for the integrity of the EU’s flagship innovation program.
1. The Facts
First Evaluation – January 5, 2024
Call: HORIZON-EIC-2023-ACCELERATOR-01
Result: 6.00 / 12 (Threshold: 9)
Document: ESR No. 101161752
Key highlights:
- Evaluators explicitly called AMPS “a groundbreaking solution integrating AdTech with environmental responsibility.”
- The work plan was “thorough and logical,” the timeline “realistic,” and the team “competent and motivated.”
- The product’s timing and market context were described as “completely right” given rising urban pollution.
- Weaknesses were limited to documentation style (AI terminology, IP structure, investor evidence).
In short: constructive feedback, clear validation, no red flags.
Second Evaluation – May 15, 2024
Call: HORIZON-EIC-2024-ACCELERATOR-02
Result: 0.00 / 12 (Threshold: 9)
Document: ESR No. 101188230
The same project was reviewed again.
The evaluators again praised:
- A clear mission and vision,
- Strong partnerships,
- A comprehensive marketing and commercialization strategy,
- A thorough, logical implementation plan.
Then the report states:
“Applicants who do not receive a GO during the remote evaluation … will receive an automatic 0 score.”
and
“An applicant may receive 2 points in Excellence, 3 in Impact and 1 in Level of risk … yet the overall score would be 0.”
That clause — inserted in 2024 — replaced the numerical scoring system with a binary gate that erases actual evaluator input.
This is where the EIC broke its own rules.
2. What the Rules Require
Under Regulation (EU) 2021/695 — Horizon Europe,
Article 22(3) stipulates that evaluation “shall be carried out in a transparent manner and shall be based on the information provided by the applicant and on clear evaluation criteria.”
Article 22(5) further guarantees that applicants may request a “review of the evaluation procedure where there is an indication of a procedural flaw, factual error or manifest error of assessment.”
The EIC Work Programme 2024, section II.2.2, repeats the same requirement:
“Each proposal shall receive a numerical score for every evaluation criterion, recorded individually by each expert evaluator.”
By assigning an automatic zero and discarding recorded criterion scores, the Commission:
- Obscured evaluator judgments — breaching the transparency clause.
- Prevented proportional scoring — breaching the proportional-evaluation principle.
- Removed the applicant’s right to meaningful review, because there are no numerical data left to challenge.
This is not an interpretation issue. It’s written black-on-white in the legal framework. The EIC’s 2024 automation effectively nullified Article 22(3) of its own founding regulation.
3. Additional Procedural Violations
a) Administrative Misidentification
Evaluator 2 flagged an inconsistency between “AddAir Corporation CEE SRL” (our legal name in the EU Participant Register) and “Airvolt Corporation CEE SRL.”
The Horizon Europe Funding & Tenders Portal automatically fills entity names from its validated database.
If the name displayed differs from our commercial brand, the evaluator must treat this as metadata, not as a basis for scoring.
Disqualifying or downgrading over such a clerical artifact contravenes EIC Evaluation Manual § 3.2, which forbids penalization for “administrative or clerical inconsistencies not affecting eligibility.”
b) Reuse of Prior Content
Sections describing TRL levels, commercialization phases, and even verbatim phrasing of risk mitigation (“thorough, logical, realistic”) were copied from the January 2024 ESR.
Resubmissions are required to be “independently re-evaluated by a new panel of experts.”
Recycling previous text without fresh assessment violates EIC Evaluation Manual § 5.4 (Independence of Evaluation Rounds).
c) Absence of Consensus Meeting
The second ESR ends with:
“The proposal did not reach the conditions necessary to be considered for the consensus meeting.”
This contradicts Work Programme 2024, § 2.3.1, which requires a consensus discussion if at least two of three evaluators give a “GO” across criteria.
Given the detailed positive comments from at least two reviewers, the failure to convene such a meeting indicates procedural negligence.
4. Why This Matters
Transparency
Applicants are entitled to know how and why they were evaluated.
Replacing numeric scores with an opaque “automatic 0” destroys auditability.
Fair competition
A binary gate favors administrative conformity over substance. Startups from smaller ecosystems — with lean teams and evolving brand structures — are hit hardest.
Public trust
EIC Accelerator funding is taxpayers’ money.
Procedural opacity erodes confidence in the fairness of EU innovation policy.
5. What We’re Doing About It
- Formal documentation.
We have archived both ESRs and a comparative procedural report that traces these inconsistencies line by line. - Right-to-review.
Under Article 22(5) we are entitled to request a Review of the Evaluation Procedure for manifest error.
The basis: automatic nullification of evaluator scores, failure to hold a consensus meeting, and administrative penalization contrary to § 3.2 of the Evaluation Manual. - Public accountability.
We believe sunlight is corrective. By publishing this analysis, we make the conversation about improving process quality, not attacking individuals. - Continuing our mission.
The substance of AMPS remains intact: validated technology, growing pilot network, real customers, and committed partners. We will finance expansion through private capital and strategic alliances rather than wait for procedural clarity.
6. Lessons for Other Founders
- Audit every form field. Treat the Participant Portal as code: one typo can override merit.
- Preserve all feedback versions. You may need them as legal evidence.
- Cite the regulation. Knowing Article 22’s wording can protect your rights.
- Do not internalize bureaucratic errors as technical failure.
7. The Bigger Picture
Europe cannot afford to let algorithmic gates replace human judgment in evaluating innovation.
If the EU expects deep-tech founders to pioneer risk, it must honor due process with equal rigor.
Otherwise, the message sent is chillingly clear: Innovation is welcome — until it confuses the system.
Conclusion
Airvolt’s AMPS was praised, scored, then zeroed.
Not because it lacked merit, but because an administrative rule allowed a computer to overwrite expert evaluation.
That is a breach of procedure, a violation of Regulation (EU) 2021/695, and a disservice to European innovation.
We will keep building.
We will keep cleaning the air.
And we will keep insisting that public innovation programs meet the same standards of accountability they demand from the innovators they judge.
Documentary sources:
- Evaluation #1: ESR 101161752 (5 Jan 2024) — HORIZON-EIC-2023-ACCELERATOR-01
- Evaluation #2: ESR 101188230 (15 May 2024) — HORIZON-EIC-2024-ACCELERATOR-02
- Regulation (EU) 2021/695, Articles 22(3) and 22(5); EIC Work Programme 2024, §§ 2.2–2.3; EIC Evaluation Manual, §§ 3.2 and 5.4.
If you are a policymaker, evaluator, or founder who believes Europe deserves better systems for judging its innovators — let’s talk.
