Enoc Martínez, Marine Environmental Researcher at the Universitat Politècnica de Catalunya, leads OBSEA, a long-term underwater observatory located off the coast of Vilanova i la Geltrú in the NW Mediterranean Sea. Since 2009, OBSEA has collected continuous multiparametric data—including temperature, salinity, currents, and acoustic measurements—providing an invaluable long-term record of marine conditions. This work is one of the ten winning projects from the second DTO-BioFlow Open Call.
This work is one of the ten winning projects from the second DTO-BioFlow Open Call. The interview was recorded during the exclusive data training workshop in Paris, held from 3–5 June 2025. The immersive event brought together experts and project teams for hands-on sessions in data transformation, quality control, metadata management, and collaborative discussions, refining methodologies and strengthening connections across the growing marine biodiversity data network.
Read the full written interview
Who are you, and what is the name of your institute?
My name is Eno Martinez, and I work for the Polytechnic University of Catalonia.
What is the focus of your work there?
We operate an underwater cable observatory—a fixed platform deployed near the coast of Vilanova, about 50 kilometres south of Barcelona. It is a multiparametric observatory that collects all kinds of data: physical parameters such as currents, waves, and temperatures, as well as biological information through underwater cameras. We have been taking underwater images since 2009, giving us a 16-year archive that we use for biological studies.
What is the goal of your current project?
Our project uses artificial intelligence to automatically detect fish species—or, more broadly, any kind of macrofauna—in the underwater images. The idea is to automatically process the data using AI algorithms and then stream the results in real time to EMODnet Biology. This would allow us to know instantly what is happening underwater and to monitor fish populations at any given moment.
How would you describe the scope of your observatory and dataset?
Even though we have a large archive of underwater images, our observatory is essentially a single point. That means it is not highly representative in terms of spatial coverage. However, we see this project as a valuable prototype for testing emerging technologies—particularly automated species identification and automated data flows—into European initiatives like EMODnet and beyond.
What is your main objective within this initiative?
The primary goal is to ensure that our data is properly connected to the major European initiatives in the biological domain. Being involved in DTOBioFlow helps us bridge artificial intelligence technologies with European data systems.
We want to automate the entire pipeline—from image acquisition to AI-based analysis, all the way to publishing the results in relevant European data portals. For us, the key benefit is ensuring our data is properly formatted, standardised, and truly usable for the scientific community.
What is your view on the Digital Twin of the Ocean (DTO)?
The Digital Twin of the Ocean is a very ambitious initiative. Whether it becomes a reality in the next 10 years is still uncertain, but the process will undoubtedly help us develop and standardise the data flows needed for large-scale scientific analytics, which is clearly the direction science is heading. We will need significant effort to ensure that data moves through the correct channels and follows the proper standards, so models and statistical methods can fully exploit the massive amount of information we are now collecting—much of which remains poorly standardised.
What is the ultimate vision for a functional digital twin?
If the digital twin becomes fully realised—although this is still far in the future—it would allow us to run “what-if” simulations.
For example:
- What would happen if there were an oil spill near our sensors?
- How would a 5°C temperature increase affect local fish populations?
- What if an entire species suddenly disappeared from the region?
The power comes from the ability to link multiple models together. That’s what makes the digital twin concept so exciting—but also extremely challenging. Only time will tell how much of it we can achieve.