1. Intrinsic, a spinout of Alphabet X, announced at the Automate conference that it is incorporating Nvidia offerings into its Flowstate robotic app platform, specifically focusing on grasping for manufacturing and fulfillment automation.
2. The collaboration with Nvidia includes ready-made universal grasping skills that can be utilized to accelerate programming processes, reduce development costs, and increase flexibility for end users.
3. Intrinsic is also working with DeepMind to tackle pose estimation, path planning, and operating multiple robots in tandem, including systems that use two arms at once for a wider range of applications.
At the Automate conference, Alphabet X spinout Intrinsic announced a collaboration with Nvidia to incorporate Isaac Manipulator into its Flowstate robotic app platform. This collaboration focuses on grasping, a key aspect of manufacturing and fulfillment automation. The goal is for robots to be able to transfer grasping skills to different settings without the need for extensive training.
Intrinsic is also working with DeepMind, another Alphabet-owned company, to crack pose estimation and path planning in automation. The system has been trained on over 130,000 objects and can determine the orientation of objects in a few seconds, crucial for picking them up. Intrinsic is also working on systems that can operate multiple robots in tandem and use two arms at once, opening up new possibilities for applications.
The collaboration with Nvidia and DeepMind enables developers to use universal grasping skills to accelerate their programming processes, reducing development costs and increasing flexibility for end users. Intrinsic is working with companies like Trumpf Machine Tools to test the system, which uses synthetic data to build sophisticated solutions for object grasping tasks in simulation and reality. The team is also focusing on efficient code generation to complete tasks using foundation models, rather than hard-coding specific grippers for specific objects.