On June 2nd 2026, Matteo Saveriano represented INVERSE in the webinar “When Robots Learn: insights into safety and trustworthiness in robotic autonomous adaptation from EU projects”, organized by the RoboSapiens project. The session brought together several EU projects for project pitches followed by a discussion on key challenges around adaptive robotics.
INVERSE at the project pitches
Matteo presented INVERSE’s motivation, objectives and challenges. INVERSE is working to solve complex manipulation tasks across different domains by synthesizing, monitoring, and executing inverse plans from direct task definitions. INVERSE’s objectives include robustly understanding a robot’s surroundings, flexibly representing knowledge, and continuously adapting that knowledge across domains.
Adaptive Robotics today – potential and barriers
Matteo presented INVERSE’s motivation, objectives and challenges. INVERSE is working to solve complex manipulation tasks across different domains by synthesizing, monitoring, and executing inverse plans from direct task definitions. INVERSE’s objectives include robustly understanding a robot’s surroundings, flexibly representing knowledge, and continuously adapting that knowledge across domains.
A particularly thought-provoking part of the discussion focused on what happens when robots begin adapting at runtime. Matteo Saveriano highlighted safety as a key challenge, noting that safety does not automatically go hand in hand with adaptation.
Another important takeaway: trust in adaptive robots grows through exposure and through safe, reliable interaction. And while the potential of adaptive robotics is enormous, there are still technical barriers to overcome before these systems can become mainstream.
Thank you to our partners
Great to see such an important exchange between EU projects working on the future of safe, trustworthy, and adaptive robotics!