Speaker
Description
This work presents a comprehensive framework for the modeling, simulation, and control of flexible robotic manipulators (FRMs), with a particular focus on remote handling applications in fusion machines. The work is motivated by the limitations of classical rigid-body assumptions when applied to long-reach, lightweight, high-degree-of-freedom robotic systems, as well as the challenges in accurately controlling and simulating these robots in constrained or hazardous environments where direct access to sensors is limited and interactions with the environment significantly affect system behavior. The proposed approach is based on a Digital Twin architecture that integrates geometrically exact flexible-body modeling with real-time, human-in-the-loop control. The simulation of FRMs relies on an efficient formulation capable of capturing large deformations while remaining suitable for real-time execution. These models are embedded within a modular software architecture that combines flexible-body simulation, vibration suppression, optimal control, collision detection based on distance-field representations, and immersive visualization through a photorealistic virtual-reality Digital Twin. The framework is validated within the NEFERTARI project, targeting remote inspection and maintenance tasks inside the RFX-mod2 fusion machine. The results demonstrate the effectiveness of the proposed approach for the design, simulation, and control of long-reach manipulators, providing enhanced situational awareness, operator support, and training capabilities in scenarios where direct sensing is limited or unavailable.