How can physical AI overcome the hurdle of scaling? A conversation with AWS's Global General Manager for Automotive and Manufacturing

What happened: At the 2026 AWS Automotive and Manufacturing Summit, AWS Global General Manager for Automotive and Manufacturing Ozgur Tohumcu discussed how AI moves from proof of concept into scaled production, and AWS's five pillars for physical AI. Key numbers: Amazon has deployed over 1 million robots across its operations network, handling millions of packages daily alongside hundreds of thousands of Why it matters: Tohumcu said data, governance and organization determine deployment speed, and that R&D is often where manufacturing and automotive customers see the earliest return.

At the 2026 AWS Automotive and Manufacturing Summit, AWS Global General Manager for Automotive and Manufacturing Ozgur Tohumcu discussed how AI moves from proof of concept into scaled production, and AWS's five pillars for physical AI.

Tohumcu said data, governance and organization determine deployment speed, and that R&D is often where manufacturing and automotive customers see the earliest return.