Collaborative Grasping

Collaborative grasping research focuses on enabling multiple robots or robots and humans to cooperatively manipulate objects, improving efficiency and robustness compared to single-agent grasping. Current efforts concentrate on developing algorithms that optimize grasping strategies based on factors like object pose uncertainty and available sensing, including reinforcement learning approaches and passivity-based decentralized control for multi-robot systems. This work is significant for advancing mobile manipulation capabilities in robotics, particularly for applications requiring complex object handling and flexible collaboration, such as warehouse automation and assistive robotics. A deeper understanding of multi-object grasping taxonomies is also emerging, informing the design of more versatile robotic hands and control strategies.