AI-REEFSHIELD

Al-driven Robotic Ecosystem Exploration Framework for Securing Habitat Integrity and Life Diversity

Project duration: 01.05.2025 - 31.10.2027

The “AI-REEFSHIELD” project is developing an AI-supported learning and robotics system for the detection and assessment of marine restoration areas. At its core is the technical development of an autonomous underwater vehicle (AUV) by the German Research Center for Artificial Intelligence (DFKI), which will be equipped with high-resolution camera technology and advanced AI methods. The goal is to map and monitor large reef areas efficiently and precisely while significantly reducing human effort, thereby advancing existing, labor-intensive monitoring approaches.

The system combines approaches from computer vision, reinforcement learning, explainable AI (XAI), and model-based learning methods. An explainable active-vision approach enables the autonomous system to selectively capture relevant environmental information and dynamically optimize its observation strategy.

The Alfred Wegener Institute (AWI) plays a central role in the project as the marine application partner and supports the validation of the developed system under real environmental conditions. The practical testing and validation of the developed technologies will take place within the Borkum Reef Ground Nature Reserve under real environmental conditions. Particular emphasis is placed on the scientific interpretation of the results in the context of the reintroduction of the European oyster and the development of marine biodiversity in the North Sea.

Through the close integration of technological development and marine field research, the project aims to create a transferable system that can, in the future, also be deployed in other sensitive marine ecosystems and contribute to sustainable environmental monitoring and restoration.

Further information on the overall project can be found from the project coordinator: DFKI project page AI-REEFSHIELD.