Prof. Ken Goldberg, PhD student Jeffrey Mahler, and researchers at the Laboratory for Automation Science and Engineering (AUTOLAB) have just created the most dexterous robot in existence.
Version 4.0 of the Dex-Net robot (more about 3.0 here) uses a 3D sensor and has two arms to grasp with, one with a gripper and the other with a suction. However, it’s superpower is its brain, Dex-Net, a cloud-based neural network that has been trained to identify 3D objects and optimally choose the best points to grasp. The software was trained by artifically simulating millions of grasps in the cloud.
“We’ve been talking about how to align our results so that we see progress,” Goldberg says. “It all depends what robot you’re using, what sensor you’re using, and—very importantly—what objects you’re using.”