The SILVANUS project, in which Eviden led T10.2 regarding exploitation and sustainability and owned the Fire Detection at the Edge component, ended successfully on May 20 after 42 months of work involving 49 consortium partners. This highlights the joint effort achieved.

To conclude the project, the EC and the coordination team organized a final review in Brussels on May 19.On behalf of Eviden, Members of the Computer Vision and the Trustworthy AI teams from R&D Spain take part in the review.

During the event, the consortium presented the outcomes of the project consisting of final impact results, post project plans and achievements accomplished.

Eviden's contribution was key to the project because the team developed the first version of an LLM for question-and-answering used in a European project.

Additionally, a fully synthetic dataset was created taking advantage of Eviden's generative AI capabilities, aligning with the internal strategy of BDS products. A detector of fire and smoke that can be further exploited within Eviden was also created.

Regarding exploitation, individual plans were gathered to identify the partners' commercialization purposes. Signed agreements from the partners supported the project's sustainability, and a market definition and analysis were also established to determine the value proposition and pathway to leadership.

Lastly, a bundle strategy was implemented to clarify future offerings tailored to end users.

Some of the main results have been as follows:

• An integrated platform involving the ten key exploitable results identified.
• Three agreements were drafted and signed by the consortium members to demonstrate their commitment: the IPR Agreement, the Exploitation Agreement, and the Letter of Intent.
• A bundle or clustering strategy. SILVANUS packages key technologies into modular bundles tailored to end-user needs, ensuring adaptability across different deployment environments.

The project achieved its expected results and has a clear pathway into exploitation and future real-world use for its solutions.