CAVECORE

Start September 1, 2025
End August 31, 2028
Project Status Ongoing
Website Website
Reference HORIZON-MSCA-2024-DN-01-01

Project overview

CAVECORE (Continuous, Automated Validation, and Evaluation of Cognitive Robots in Open-Ended Environments) is a pioneering research and training project focused on advancing cognitive robotics systems that can autonomously interact, learn, and adapt in complex, real-world environments. The project introduces the concept of “evaluability-by-design”, embedding validation and testing into every stage of robot development to ensure reliability, safety, and trustworthiness. Through an interdisciplinary approach combining robotics, cognitive science, AI metrology, and automated testing, CAVECORE establishes a new standard for evaluating cognitive robots in compliance with the EU AI Act. By developing tools for continuous self-evaluation and training a new generation of researchers in replicable and unbiased methods, CAVECORE paves the way for transparent, ethical, and high-performance robotic systems ready for real-world deployment in sectors such as healthcare, logistics, and scientific exploration.

€ k
project funding
9
partners involved
6
countries involved
CAVECORE

Our Role

PAL Robotics plays a central role in the CAVECORE project as an industrial and training partner, leading the development and validation of methods for evaluating shared models in human-robot understanding. As host and supervisor of a doctoral candidate, PAL Robotics contributes to defining best practices for trustworthy and replicable robot evaluation. We also lead the establishment of the Industrial Forum Group (IFG) and organize key workshops and training events to bridge academic research and industrial application.

Key responsibilities

  • Lead task  for the progressive evaluation of shared models for Human-Robot understanding
  • Coordinate  the establishment of the Industrial Forum Group (IFG)
  • Supervise a doctoral candidate from month 6 to 42 and host six additional candidates for secondments across three years
  • Organize and host workshops and training events
  • Guide the creation and monitoring of Career Development Plans (CDPs) for multiple Doctoral Candidates

Partners