ENRESA-CSIC 2026 Agreement: €545,200 for AI in Radioactive Waste

Key data

The Spanish nuclear sector takes a significant step in waste management: ENRESA will allocate up to 545,200 euros to the CARINA project, a collaboration agreement with CSIC formalized on August 5, 2026 and published in the BOE on August 13 (reference BOE-A-2026-17700). The objective is to develop innovative techniques for characterizing medium, low and very low activity radioactive waste, combining neutron-gamma imaging with computer vision and artificial intelligence.

This is not a generic public subsidy: it is an operational agreement with specific timelines, amounts and technical objectives, which has direct implications for the spent nuclear fuel management chain and decommissioning processes of nuclear facilities in Spain.

What does this regulation establish?

The agreement formalizes collaboration between ENRESA (National Company for Radioactive Waste, S.A., S.M.E.) and CSIC (Spanish National Research Council), with the latter acting through IFIC (Institute of Corpuscular Physics), for the development of the project called CARINA: "Innovative characterization of LMLVA through neutron-gamma imaging assisted by computer vision and AI techniques".

The acronym LMLVA refers to Low, Medium and Very Low Activity Waste, which constitutes the bulk volume of waste generated in Spanish nuclear facilities. Neutron-gamma imaging technique allows identifying and quantifying the radioactive content of containers without opening them, and the incorporation of AI and computer imaging seeks to automate and improve the precision of that process.

Economic and operational impact

The direct economic impact of this agreement falls on ENRESA, which assumes the full cost of up to 545,200 euros within its 9th R&D Plan 2024-2028. For CSIC-IFIC, it represents the acquisition of funding for applied research in a highly specialized area.

From an operational perspective, the development of more precise and automated characterization techniques has far-reaching consequences:

• Reduction of management costs: better waste characterization allows optimizing its classification, storage and treatment, avoiding additional costs from incorrect classifications.
• Safety in decommissioning: the decommissioning of Spanish nuclear plants requires characterizing large volumes of waste. More precise technology reduces risks and timelines.
• Reinforced regulatory compliance: Royal Decree 102/2014 requires responsible and documented management of radioactive waste. Better characterization tools facilitate this compliance.
• Extended time horizon: with the possibility of extension up to 10 years, the project can accompany the complete decommissioning cycle of several facilities.

Who does it affect?

• ENRESA: as the funding entity and direct beneficiary of the developed technology.
• CSIC-IFIC: as the executing entity of the research and recipient of the funding.
• Operators of Spanish nuclear plants: that generate the LMLVA waste that will be subject to the new characterization methodology.
• Companies in the nuclear decommissioning sector: that may benefit from the techniques developed in their processes.
• Technology suppliers of AI and industrial imaging: with potential for collaboration or technology transfer derived from the project.
• Nuclear regulatory authorities (CSN): that supervise compliance with Royal Decree 102/2014 and may validate the new characterization methods.

Practical example

Imagine a nuclear services company contracted by ENRESA for the decommissioning of a facility. Currently, characterizing a batch of LMLVA waste containers requires manual processes or conventional techniques that consume time and generate uncertainty in classification.

With the technology that the CARINA project will develop, that same batch could be analyzed using neutron-gamma imaging combined with AI algorithms, obtaining an automated classification, faster and with less margin for error. This translates to:

• Reduction of characterization time per batch.
• Lower risk of subsequent reclassification (which implies additional costs and regulatory delays).
• More robust documentation for compliance with Royal Decree 102/2014.

The agreement has an initial duration of 5 years and can be extended up to 10 years, which means that results could be available for industrial application before 2031, with the possibility of continued use until 2036.

What should companies do now?

1. Identify if your activity generates or manages LMLVA waste: if you operate in the nuclear or decommissioning sector, this project may affect your characterization processes in the 2026-2031 horizon.
2. Review compliance with Royal Decree 102/2014: the agreement is framed within this regulation. Ensure that your radioactive waste management procedures are up to date.
3. Follow the evolution of the CARINA project: the R&D results of ENRESA-CSIC may lead to new standards or technical requirements for characterization that affect your operations.
4. Evaluate opportunities for technological collaboration: companies supplying AI, computer vision or neutron-gamma imaging equipment can explore ways to participate or transfer technology with CSIC-IFIC.
5. Consult the CSN (Nuclear Safety Council): if you are a nuclear operator, verify whether the new characterization methods that emerge from the project will require updating your approved licenses or protocols.

Official source

Consult complete regulation in official source

Notice: This article is for informational purposes only and does not constitute legal advice. For specific decisions, consult a qualified professional. Source: https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-17700