CoreSense

CoreSense is a Horizon Europe research and innovation action that develops a cognitive architecture for autonomous robots, grounded in formal theories of understanding and awareness. The Universidad Politécnica de Madrid coordinates it, and CVAR-UPM builds its drone demonstrator on Aerostack2.

• Programme. Horizon Europe, call HORIZON-CL4-2021-DIGITAL-EMERGING-01
• Grant agreement. 101070254
• Dates. October 2022 to September 2026
• Coordinator. Universidad Politécnica de Madrid
• Links. coresense.eu and CORDIS

The project

Robots now work in open environments that their designers can't fully foresee. They often don't understand what is going on around them or their own role in it, which makes them hard to rely on and hard to trust. CoreSense answers this with a theory of understanding, a theory of awareness, reusable software that applies both to real robots, and three demonstrations of what they bring. The demonstrations make drone teams more resilient, manufacturing robots more flexible, and social robots better aligned with the people around them.

Aerostack2 in CoreSense

The drone demonstration is the inspection testbed, where a swarm of drones inspects an array of solar panels. The drones split the area through an auction, fly their coverage lanes, and take over the waypoints of a drone that fails mid-mission. Aerostack2 is the runtime of the whole testbed, and its simulation runs entirely in the Aerostack2 multirotor simulator.

• TB2_Panel_Inspection_Simulation holds the simulated inspection missions, written with the Aerostack2 Python API.
• as2_knowledge_graph is the knowledge graph behind the knowledge-driven missions listed below.
• The simulated dataset and the ROS bag dataset record the testbed's missions.

Publications

• G. GP-Lenza, C. D. R. Pita-Romero, M. Fernandez-Cortizas and P. Campoy. A Methodology for Designing Knowledge-Driven Missions for Robots. arXiv:2601.20797, 2026.
• C. D. R. Pita-Romero, P. Arias-Perez, M. Fernandez-Cortizas, R. Perez-Segui and P. Campoy. Flocking Behavior for Dynamic and Complex Swarm Structures. arXiv:2601.21772, 2026.
• G. GP-Lenza, M. Fernandez-Cortizas, M. Molina and P. Campoy. Distributed Coordination for Resilient Multi-UAV Remote Sensing: A Photovoltaic Inspection Case Study. arXiv:2607.24482, 2026.