[REVISION HISTORY]

Edge AI and Physical AI expansion across APAC

Updated 3 times since CLSTR started tracking revisions of this situation.

What changed

2026-08-17 01:08 UTC → 2026-08-21 09:21 UTC · added ~~removed~~

Throughout mid-2026, a series of strategic partnerships accelerated the deployment of Edge AI and physical AI across the APAC region, with a heavy focus on Japan and industrial automation. In late June, Aptiv and Nvidia expanded their collaboration to accelerate production-ready edge AI solutions, specifically targeting the Nvidia Jetson platform. This was followed in July by DEEPX and Avnet launching a 15-market distribution network across APAC. Kyndryl also expanded its edge computing footprint through a partnership with Aptiv to deploy Wind River software technologies. Nvidia’s influence deepened significantly in Japan through partnerships with Toyota and the expansion of its Cosmos Coalition with partners including Honda R&D and SoftBank. This momentum led to multibillion-dollar contracts for Japan’s Noetra national AI infrastructure. Industrial integration continued with Tokyo Electron expanding its tie-up with Nvidia to embed the Agent Toolkit and Isaac robotics into its Epsira™ suite. This expansion utilizes Nvidia Omniverse simulation libraries to create digital twins, aiming to improve production efficiency and reduce downtime through AI-driven troubleshooting. In mid-August, the push for sovereign physical AI was formalized as Fujitsu announced a strategic collaboration with robotics leaders Fanuc, Yaskawa Electric, and Kawasaki Heavy Industries. Utilizing Nvidia technology, the partnership aims to build a coordinated control and task planning platform to connect digital systems with physical robot control. This initiative seeks to address labor shortages in manufacturing, logistics, and medical services. As part of this broader technological shift, Fujitsu plans to invest approximately 3 trillion yen over the next decade into new business fields, including physical AI. Concurrently, Nvidia introduced Cosmos 3 Edge, a 4B parameter model optimized for on-device robot control on Nvidia Jetson Thor hardware. The model uses pre-trained physical-world data to predict robot actions within closed-loop simulations.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.