Explainable AI for robots


Cognitive robots are augmenting their autonomy, enabling them to deployments in increasingly open-ended environments. This offers enormous possibilities for improvements in human economy and wellbeing. However, it also poses strong risks that are difficult to assess and control by humans. The trend towards increased autonomy conveys augmented concerns on reliability, resilience, and trust for autonomous robots in open worlds. The essence of this issue can be traced to robots suffering from a lack of understanding of what is going on and a lack of awareness of their role in it. To adress these issues, we will develop a cognitive architecture for autonomous robots based on a formal concept of understanding, supporting value-oriented situation understanding and self-awareness to improve robot flexibility, resilience and explainability.


Autonomous robots do not fully understand their open environments, their complex missions, their intricate realizations, and the unexpected events that affect their performance. This is a subject that artificial intelligence approaches based on machine learning are not addressing well. Therefore, an improvement in the capability to understand of autonomous robots is needed. This project tries to provide a solution to this need in the form of:

  • A theory of understanding
  • A theory of awareness
  • Reusable software assets to apply these theories in real robots
  • Three demonstrations of its capability.

This will lead to increasing the resilience of drone teams, improving flexibility of manufacturing robots, and augment human alignment of social robots.

The project has three development threads – theory, technology, application – with three associated objectives related to project products, and one impact thread with an objective related to the robot software community. The four threads are:

  • A theoretical thread: formulation of the theory of awareness based on a systemic concept of understanding.
  • A technological thread: implementation of a reference architecture and engineering toolbox for aware autonomous robots.
  • An application thread: use of the theory and the technical assets in the construction of autonomous robots, with increased capabilities to work without/with limited supervision; as well as the next generation of interactive robots, with greatly improved intuitive, safe and efficient cognitive, social and physical capabilities, to assist humans.
  • An impact thread: creation of an open-source community in the ROS ecosystem around the architectural software developed in the project.

SAM XL is active in the application thread and will develop a manufacturing testbed for the quality inspection of the manufacturing of large parts combined with mobile manipulators. Robotic technologies inserted in manufacturing cells will address the problem of flexible adaptation to changing production conditions relating to the locations of the robot, the tools needed, the need of learning new operations or the unexpected events related to finding defects in manufactured parts.


TU Delft, Universidad Politecnica de Madrid, Fraunhofer IPA, PAL Robotics, Universidad Rey Juan Carlos, Irish Manufacturing Research.


The Coresense project is funded by the EC Horizon Europe programme though grant HE #101070254 inside the HORIZON-CL4-2021-DIGITAL-EMERGING-01-11 topic. 

More information

Find more information at https://coresense.eu/ 

Duration: 2020 – 2024