xArm Handovers: Facilitating human to robot handoffs

Group: Neel Sortur and Tom Wilson

In this project, we trained an agent to dynamically hand over an object to a human. Using deep reinforcement learning, we train the arm to identify an object placed in front of it. It repeatedly chooses a small joint movement that will result in a position closer to a grasp, eventually getting close enough to grasp the object from a user’s hand.

”

For information about our methods and results, please visit the official project website.

Check out the code on GitHub.