Robot DexNet 2.0 - 99% precision grasping

Grabbing the awkwardly shaped items that people pick up in their day-to-day lives is a slippery task for robots. Irregularly shaped items are easy for people to grab and pick up, but robots struggle with knowing where to apply a grip. In a significant step toward overcoming this problem, researchers have a built a robot, called DexNet 2.0, that can pick up and move unfamiliar, real-world objects with a 99% success rate.
DexNet 2.0 gained its highly accurate dexterity through deep learning. The researchers built a vast database of three-dimensional shapes that a neural network uses to learn grasps that will pick up and move objects with irregular shapes (News Item UC Berkeley, 26 May 2017).
Click here for the news item.

This news item is also included in our monthly overview, the NVC Members-only Update. If you have any questions, please contact us: info@nvc.nl, +31-(0)182-512411.