BMW and Nvidia train logistics robots with AI to move car parts


BMW announced Wednesday it is joining forces with chip maker Nvidia to create faster and smarter logistics robots.

In a blog, BMW announced it has been using a variety of logistics robots that are “proving vital in the supply chain as they can carry various products and parts from one bay to another, without anyone watching them.” The technology will be rolled out to every BMW plant globally.

BMW has logistics robots and smart transport robots (STR) that were developed in-house with AI modules. The robots use the AI to recognize people and objects like forklifts as they navigate the floor.  They can even calculate alternative routes in milliseconds, BMW said.

Sponsored by Digi-Key

Boost productivity, safety and data reliability by using Excelon™ F-RAM for data logging in mission-critical applications

Cypress’ next-generation Excelon™ F-RAM family offers up to 8Mbit of high-speed nonvolatile data logging with virtually unlimited endurance and instant data capture required for factory automation and industrial IoT systems.

BMW used Nvidia’s Isaac robotics software to develop functions of perception, segmentation, pose estimation and human pose estimation to make the robots more agile and to avoid collisions.  The robots are even able to render tracings of machine parts, which are used to train neural networks on Nvidia DGX servers.

Jurgen Maidl, head of logistics at BMW Group’s Production Network, called the use of AI and visualization technologies for logistics “revolutionary.”

Nvidia has been active in AI development for at least five years and has partnerships with dozens of carmakers who are also incorporating the technology inside of automated vehicles and future self-driving vehicles.  Nvidia is expected to describe its AI innovations in its GTC 2020 conference being held online for 45,000 developers this week.

Nvidia has been active in data center AI as well and announced the completion of its $7 billion Mellanox Technologies acquisition on April 27, more than a year after it was first announced.

Nvidia gained access to Mellanox’s high-performance networking technology used in data centers, which will help expand Nvidia’s existing role in high performance computing and data centers. 

Nvidia CEO Jensen Huang said in a statement that Mellanox will give Nvidia “end-to-end technologies from AI computing to networking, full stack offering from processors to software and significant scale to advance next-generation data centers. “

Mellanox CEO Eyal Walman said the technologies from both companies will bring compute and storage solutions “wherever they are required.”

Following the purchase, Nvidia stock rose to a record high on Tuesday May 12, hitting 325.40, but fell to 311.20 at market close on Wednesday.  Needham analyst Ravindra Gill said the Mellanox purchase will be “highly accretive” with increased spending by hyperscale cloud data center providers.

 Nvidia also plans to buy Cumulus Networks but hasn’t disclosed the value of the deal.

RELATED: Nvidia unveils Orin chip for robots and AVs

 



Source link Google News

Share