Five Boxes Hold Two Hundred Million

The budget was divided before the call opened. EUR 220 million across five Challenges, EUR 20 to 50 million per Challenge, spread across work programme pages 90 to 99. That is not a menu. It is five rooms, and each room has its own door. The instruction sits in one line of the overview: if the project does not fit any Challenge, apply to EIC Accelerator Open instead. That line is the article. The topic is chosen before the technology is, by somebody writing a work programme months earlier.

Challenge 1 wants advanced materials for energy generation and storage. Its background is a number: renewables reached 47 per cent of EU electricity generation in 2024, against 34 per cent in 2019, and most of the metals and rare earths behind that rise are not sourced or processed in Europe. The objectives name design, synthesis, characterisation, scale-up, a cut in critical raw materials, and a life-cycle analysis carrying both cost and social impact. Challenge 2 is fusion. Alternative concepts, materials that take neutron flux, tritium production, high-repetition lasers, magnets, digital twins, inertial fusion targets. Commercialisation is expected around 2040, and part of the budget is transferred from Euratom. Challenge 3 is soil. 60 per cent of EU soils are unhealthy, and across 14 European countries the land sensitive to desertification nearly doubled to 400,000 square kilometres between 2008 and 2017. Three focus areas: bioremediation, soil microbiome management, and renewable fertilisers drawn from waste streams and biomass. Challenge 4 is the raw materials chain behind the Critical Raw Materials Act of 2024 and its 2030 benchmarks for domestic capacity. Exploration, extraction and metallurgy, then recycling from end-of-life products. Deep-sea mining is out of scope, and stated as out of scope. Challenge 5 is climate adaptation: urban heat, climate-smart agriculture, water scarcity, flood and coastal protection. It connects to a mission that already has more than 400 regional and local authorities signed up to a 2030 resilience target.

Open and Challenge are not two doors into one room. A Challenge dossier is scored on criterion 2.4, contribution to the Challenge's own objectives, with metrics attached to the claim. For STEP Challenges, criterion 2.5 asks for a route into the EU internal market with significant economic potential, or a reduction of a strategic dependency. Part B carries a Challenge-specific section spelling out the fit. So the same document is read by the same evaluators against a different question. In Open, the question is whether the technology is excellent. In a Challenge, it is whether this technology is the answer to a published question. Those are not the same exam, and they do not reward the same sentences.

Challenge 4 carries a restriction with nothing to do with the engineering. Grant-only applications need an Ownership Control Declaration, and the assessment decides whether the company is controlled by a non-associated third country. The same declaration follows Open proposals that fall inside that Challenge's scope. Associated Countries carry their own load: they have to show a contribution to the EU internal market or to a strategic dependency, not merely a good project. An engineer reading the topic list will look for the technology. The first filter is corporate structure.

The topic is chosen before the technology is. The dossier arrives already filed against a question somebody else wrote months earlier.

The comparison matrix puts four of the five Challenges at TRL 5 to 8 and fusion at 4 to 6, with a life-cycle analysis required in three of them. A fit assessment sitting in my own files rates one waste-to-energy technology in the vault - gasification of municipal and agricultural residue - MEDIUM-HIGH on Challenge 5, through agricultural residue and landfill methane, MEDIUM on Challenge 1, through the catalyst and the syngas, and HIGH on Open. Three ratings for one machine, decided by which box it is filed in. Nothing about the reactor changed between the three readings. The probability that a project filed into a Challenge it does not match is then read as the project it actually is is not zero. Most people price it at zero and file anyway.

Sixteen years of field work across Nigeria, Serbia, Kazan and Fukushima taught me one habit that has nothing to do with technology: read the form before the work. Everywhere I have built, the shape of the application decided which of my projects got seen, and the ones that went unseen were rarely the weakest ones. A topic is a door. You cannot walk through the one you did not choose.