www.industry-asia-pacific.com
17
'26
Written on Modified on
Integration of Physical Artificial Intelligence Platforms in Japanese Industry
A consortium of Japanese enterprises is deploying NVIDIA Cosmos and Isaac technologies to optimize production, logistics, and autonomous systems.
www.nvidia.com
Japanese industrial and technology companies have partnered with NVIDIA to deploy physical artificial intelligence systems across robotics, manufacturing, and digital infrastructure. This cooperation integrates the NVIDIA Cosmos, Isaac, Metropolis, and Jetson platforms with the partners' existing operating systems to enable local data processing and real-time decision-making.
Operational Challenges and Cooperation Objectives
Scaling industrial automation systems in complex environments requires high precision, low latency, and the ability of machines to adapt to dynamic surroundings. Traditional programming methods for robots and vision systems are time-consuming and difficult to adapt to new conditions. This integration required combining NVIDIA's accelerated computing and AI model expertise with the automation experience of Japanese industry leaders, including FANUC, Yaskawa Electric, Kawasaki Heavy Industries, and Fujitsu, to establish a standardized operational environment.
Technical Architecture and Division of Responsibilities
The deployment is built on the Cosmos 3 Edge model, featuring 4 billion parameters optimized for local execution on NVIDIA Jetson edge computers, including the T2000 and T3000 modules. This model allows for local environmental analysis and the generation of machine motion trajectories without requiring constant data transmission to the cloud.
The division of responsibilities across the implementations includes:
- Fujitsu, FANUC, Yaskawa Electric, and Kawasaki Heavy Industries: Developing a shared control platform that integrates NVIDIA software with physical execution systems. This platform uses Isaac robotics libraries and the Newton physics engine to manage the design cycle from simulation to real-world deployment.
- SoftBank Corp.: Constructing AI-RAN edge computing and connectivity infrastructure using the NVIDIA AI Aerial platform to ensure stable data transfer for physical devices.
- Mujin and TRON K.K.: Integrating physical models into their proprietary control systems, such as MujinOS, to automate assembly, sorting, and quality inspection tasks.
Application Areas and Pilot Deployments
Test and operational deployments are being executed across key economic sectors:
Test and operational deployments are being executed across key economic sectors:
- Precision Agriculture: Kubota is exploring the use of Cosmos models to control autonomous agricultural machinery.
- Smart Building Systems and Infrastructure: Hitachi and Shimizu Corporation are deploying Metropolis vision systems to manage safety on construction sites and optimize energy consumption in commercial properties.
- Assistive Robotics: Enactic is adapting the Isaac GR00T robotics model for semi-humanoid elder-care robots.
By utilizing open Metropolis libraries, developers can configure and deploy local robot action models in less than a day, reducing production line downtime during task reconfiguration.
Edited by Evgeny Churilov, Induportals Media - Adapted by AI.
www.nvidia.com
Edited by Evgeny Churilov, Induportals Media - Adapted by AI.
www.nvidia.com
