SHEREC

SHEREC is a Horizon Europe innovation action that brings robots, data and AI into ship recycling. CVAR-UPM is a partner and develops the localization and navigation of its inspection drone on Aerostack2.

• Programme. Horizon Europe, call HORIZON-CL4-2023-HUMAN-01-CNECT
• Grant agreement. 101136056
• Dates. January 2024 to December 2027
• Coordinator. HIDROPAR, Türkiye
• Links. sherecproject.com and CORDIS

The project

Ship recycling takes ships apart so their materials can be reused, which makes it part of Europe's circular economy. It is also one of the most dangerous jobs in the world, with high rates of injury and exposure to toxic waste. SHEREC sets out to make it safer and cleaner in three ways.

1. An AI-powered drone inspects the inside of the ship and finds hazardous materials.
2. A digital twin of the ship and AI planning produce the recycling plan automatically.
3. Two mobile robots, autonomous or teleoperated, carry out the cutting and paint removal.

Aerostack2 in SHEREC

The inspection drone flies inside ships, where there is no GPS. CVAR-UPM leads its semantic SLAM, which keeps the drone's position free of drift by recognising walls, rooms, doors and equipment as landmarks.

• sherec_semantic_slam is deliverable D2.4 of the project. It runs onboard as an Aerostack2 node and publishes the corrected pose that navigation uses, along with a semantic map of the ship's interior.
• project_sherec_navigation is an example of the Aerostack2 navigation developed for the project, which runs in simulation with Docker.

Publications

• D. Perez-Saura, M. Fernandez-Cortizas, A. J. Gaona and P. Campoy. Dual Pose-Graph Semantic Localization for Vision-Based Autonomous Drone Racing. arXiv:2604.15168, 2026.
• A. J. Gaona, D. Perez-Saura, F. J. Anguita and P. Campoy. Flight-Ready LiDAR-Inertial Odometry for Embedded Drone Platforms. arXiv:2607.22145, 2026.