Teaching robots new skills is not hard but they still have challenges when trying to emulate some of the things that come naturally to us. For example, the idea of grasping both stationary and moving objects in one motion is not always easy to pull off. Finger Flow is a reactive motion generator that uses feedback from an onboard camera and position feedback from fingers and arms to grasp stationary and moving objects.
For this project, neoDavid was used, which is a humanoid robot with a multi-fingered hand. The above video shows it in action.
[HT: X. Meng, H. Maurenbrecher, A. Albu-Schäffer, and M. Keppler. “Finger flow: Reactive reach-while-grasp generation for robotic arms with multi-fingered hands.” Robotics and Autonomous Systems (2026): Volume 195, 105222]