A Horizon Europe topic anticipates how AI will reach industry

European artificial intelligence is entering a more demanding stage. The European Commission is no longer seeking only to fund new technological capabilities, but also to accelerate their arrival in factories and SMEs. This evolution is beginning to take concrete shape in some Horizon Europe funding lines. One of the clearest examples is the topic HORIZON-CL4-2027-04-DIGITAL-EMERGING-05, dedicated to the integration and adoption of robotics based on artificial intelligence. The call continues to support the development of new algorithms, sensors and robotic systems, but shifts the centre of gravity towards the integration of already available technologies, their validation in industrial environments and faster adoption by companies.

“This topic provides a very clear view of where Brussels is heading,” explains Luis Javier Pérez, consultant in the Digital area at Zabala Innovation. “Recent documents such as the State of the Digital Decade 2026 and the Apply AI Strategy were already pointing in that direction. The difference is that these priorities are now beginning to translate into concrete requirements within calls,” he adds. Technological development will remain necessary. The European Commission, however, is now asking for technology to be connected to specific industrial needs, integrated with other existing resources and able to generate reusable results.

From isolated project to shared ecosystem

In this way, the topic encourages applicants to make use of European assets such as AI-on-Demand resources and the EuroCORE repository, as well as AI models, datasets and benchmarks generated in previous initiatives. “This shift will result in less isolated projects,” predicts Pérez, based on the observation that “if every project rebuilds the same pieces, progress slows down; if it starts from previous results and contributes new components to the ecosystem, the impact is multiplied”.

Companies applying for funding for these projects will need to explain which existing resources will be used, which components other actors will be able to reuse and how adoption by industrial users will be facilitated. “The value of a proposal will depend not only on what it is able to develop, but also on how it integrates into the European ecosystem and what it leaves available for those who come afterwards,” the expert notes.

From the laboratory to industry

In this respect, industrial robotics based on AI offers a good field in which to observe this evolution. Europe has scientific knowledge, technology centres, manufacturers, integrators and user companies, but large-scale adoption continues to face barriers. Many solutions work in the laboratory, but are difficult to deploy in production environments with constraints relating to safety, costs, interoperability, maintenance and skills.

For this reason, projects will need to demonstrate that their solutions can be tested in industrial cases, connected to existing infrastructures and respond to real needs in productive sectors. Instruments such as the Testing and Experimentation Facilities (TEFs), the European Digital Innovation Hubs (EDIHs) and the AI Factories can play a complementary role: validating technologies, bringing them closer to SMEs and reducing the distance between the laboratory and the market.

A new way of building consortia

This approach also changes the way consortia are understood. European collaboration is no longer only about bringing together partners from several countries. It also requires better connections with platforms, standards, repositories, experimentation infrastructures and technical communities in the European AI ecosystem. “The companies that position themselves best will be those that can explain not only which industrial problem they want to solve, but also how they will make use of what Europe has already funded and how they will turn their results into something useful for others,” Pérez points out in this regard.

In his words, “not all future Horizon Europe calls will necessarily follow the same model, as each topic responds to different objectives and sectors”. Nevertheless, “the trend points to an increasingly visible criterion in European innovation policy: reducing duplication, increasing interoperability and turning R&D results into solutions that companies can use,” he concludes.