Exhibitors & Products
Events

And robotics is now also attracting the really big technology companies: Siemens was there, Google, Bosch anyway and NVIDIA seized the opportunity and brought a lot with them to Denmark. The Americans used the conference to launch a product offensive. The developers announced generative AI tools, simulations and perception workflows for ROS. The basis is the Jetson platform. Edge AI and robotics are tempting markets for NVIDIA and Co.

Generative AI enables robots to perceive and understand the context of their environment, communicate naturally with humans and make adaptive decisions independently, promises NVIDIA. This is what ReMEmbR is for. It is based on ROS 2 and uses generative AI to improve the logical thinking and actions of robots. It combines large language models (LLMs), visual language models (VLMs) and retrieval-based generation to enable robots to build and retrieve long-term semantic memories and improve their ability to navigate and interact with their environment.

The ROS 2 robots with voice control utilize the NVIDIA Riva ASR-TTS service to enable robots to understand and respond to spoken commands. NASA's Jet Propulsion Laboratory independently demonstrated ROSA, an AI-powered agent for ROS running on its Nebula SPOT robot and the NVIDIA Nova Carter robot in NVIDIA Isaac Sim.

At ROSCon, Canonical will demonstrate NanoOWL, a zero-shot object recognition model running on the NVIDIA Jetson Orin Nano System-on-Module. It enables robots to identify a variety of objects in real time without having to rely on predefined categories.

And the “hosts” of Universal Robots also rely on NVIDIA. They presented their UR AI Accelerator - a hardware and software toolkit designed to further advance the development of AI-supported cobot applications.