OceanX and IMAR Complete Digital Twin of Cabo Verde’s Deep Sea to Guide Global Ocean Protection

OceanX and Cabo Verde’s Instituto do Mar (IMAR) have concluded a high-impact deep-sea mission that could reshape how nations protect remote marine ecosystems. Over ten days of continuous operations, scientists aboard the OceanXplorer used eDNA, imaging, oceanography, and AI to model the biodiversity and habitat of the Nola Seamount complex, an effort designed to support national strategy and global marine protection targets. 

The expedition, Cabo Verde: Digital Deep, marks a leap forward in how deep-sea environments are studied and protected. It is among the first missions in the region to link AI-powered prediction tools with direct seafloor observations and molecular data, giving researchers and policymakers a powerful new method for identifying areas of ecological importance. 

The predictive model developed during the mission allows scientists to anticipate where vulnerable marine ecosystems (VMEs) may exist across unexplored areas. This approach aims to help identify priority conservation areas across similar seafloor habitats, including those that have not yet been physically surveyed. The technology is designed for use beyond Cabo Verde, offering a scalable framework for countries working to meet global 30x30 marine protection targets. 

“Cabo Verde has given us a unique opportunity to push the boundaries of how science and technology can support marine protection,” said Dr. Vincent Pieribone, Co-CEO and Chief Science Officer of OceanX. “By working together to build a digital twin of this ecosystem, we’re not only training AI models to predict activity on the ocean floor, we’re also creating a tool that can help Cabo Verde’s scientists and policymakers make informed decisions about their ocean. This collaboration shows what’s possible when we align shared priorities with cutting-edge science.” 

This is one of the first deep-sea expeditions to use AI-driven habitat prediction at this scale. 

“This collaboration brings together global tools and local expertise to deliver something greater than either could achieve alone,” said Dr. Yara Rodrigues, joint mission coordinator at IMAR. “The data we’re collecting will help build a clearer, more dynamic picture of our ocean. One that informs national policy, strengthens regional science, and supports better protection of our marine environment over time.” 

Mission Outcomes 

During the ten-day expedition, teams: 

• Deployed SeaSwipe, an AI platform developed in-house by OceanX, to assist with real-time species classification during ROV dives. The tool was used to analyze 252,867 image patches, retrain models 351 times, and generate 1.7 million predictions with 69 percent accuracy. 

• Processed 12,733 liters of seawater using an in-situ eDNA filtration system mounted on the ROV, generating 598 samples to map biodiversity 

• Conducted 25 ROV dives, 11 manned submersible dives, 10 CTD casts, and over 1,500 kilometers of EK80 and ADCP acoustic survey lines 

• Collected approximately 300 oceanographic samples and more than 50 terabytes of scientific data 

• Logged 10,787 visual annotations and over 60 hours of analyzed footage to classify deep-sea species in real time 

• Re-mapped the seamount complex with multibeam sonar to enhance habitat resolution and support high-fidelity ecological modeling 

The resulting model integrates species presence data from both eDNA and visual sources, enabling a more complete picture of the ecosystem and helping identify habitat conditions that support vulnerable taxa like deep-sea corals and sponges. These insights contribute directly to Cabo Verde’s national ocean strategy and provide a framework for similar efforts globally. 

A Scalable Model for Global Impact 

The Digital Deep expedition builds on earlier collaboration between OceanX and IMAR during the Around Africa 2025 campaign. This latest mission demonstrates how focused, cross-sector scientific partnerships can deliver practical tools for marine governance. As governments and conservationists work to protect 30 percent of the ocean by 2030, the predictive methodology developed in Cabo Verde offers a way to accelerate action with data that is rigorous, accessible, and directly applicable. The insights from Cabo Verde could change how governments around the world approach deep-sea conservation, especially in regions where data is limited and access is difficult.