NVIDIA: scaling physical AI demands safety at every layer, from AVs to robots

nordicinst · x · 2026-09-22

An NVIDIA blog post by Riccardo Mariani argues that safety for physical AI must span hardware, software, AI behavior, operating environment, and the full deployment lifecycle — not a one-time pre-deployment check. ABI Research projects 49 million level 3-5 autonomous vehicles by 2035, and Omdia estimates 60 million industrial robots deployed between 2026 and 2035. As these machines share roads, factories, and warehouses with people, manufacturers, regulators, and insurers need evidence they operate safely without human intervention. The post outlines four shifts defining new safety standards and a full-stack approach from hardware reliability to fail-safe fallbacks.

Original post →

More from Embodied

Embodied channel →