Innovation Action (IA)

An innovation action builds, demonstrates, pilots or replicates something that already works in principle. It sits at technology readiness levels 5 to 8: prototyping, testing in a relevant or operational environment, first-of-a-kind demonstration, market replication. It is not where you find out whether an idea works. It is where you prove it works at scale.

The funding rate is 70% of eligible costs for profit-making organisations and 100% for non-profit legal entities, and both rates apply inside the same project. A university partner is reimbursed in full while the company beside it covers 30% of its own costs from its own resources. That asymmetry has to be planned for at budgeting time, because it decides which partner can afford which work package.

The 30% is real money leaving your company, not a paper contribution. For a partner with a €1 million budget share that is €300,000 of its own cash across the project's life, and it is the most common reason a keen SME withdraws between proposal and grant agreement.

Consortium requirements match the RIA: three independent legal entities from three different Member States or Associated Countries. Evaluation uses the same three criteria and the same thresholds, but where two proposals tie, Impact is decisive for an IA (the mirror of the RIA, where Excellence breaks the tie).

Deliverables are heavier than in a RIA. Demonstrators, pilot lines and validated systems are expensive, slower than planned and dependent on suppliers, and the reviewers will ask to see the thing. A work plan that treats a physical demonstrator with the same contingency as a report is optimistic.