The era of physical artificial intelligence (AI) has officially begun. At CES 2026, humanoid robots, autonomous driving, digital twins, and wearable AI emerged as key themes. In March, at NVIDIA's GTC 2026, CEO Jensen Huang declared a "big bang of physical AI," confirming the transition to real-world applications. The infrastructure is now in place for AI to move from virtual training environments to actual factories and logistics sites.
Despite appearing as distinct technologies, all these advancements share a common thread: the ability to perceive and interpret the physical world. AI operating in the real world must be supported by a precise understanding of complex conditions on the ground.
To effectively utilize spatial recognition technology in industrial settings, a connection point between workers and AI is essential. Smart glasses have emerged as a focal point. In 2026, major global tech companies began seriously exploring smart glasses as an AI interface for industrial applications.
At CES 2026, Siemens announced a collaboration to integrate industrial AI into Meta Ray-Ban AI glasses. When factory workers wear these glasses, they receive real-time voice guidance and task feedback hands-free. Siemens and NVIDIA further declared their intention to jointly develop an "industrial AI operating system," outlining a roadmap to realize the world's first fully AI-based adaptive factory starting at the Erlangen plant in Germany. Google is also expanding its smart glasses ecosystem targeting the enterprise market through its Android XR platform.
Market research firm Omdia predicts that shipments of AI glasses will surpass 15 million units this year and reach 35 million by 2030. While it may take time for robots to fully replace factories, equipping workers with AI is feasible right now.
For smart glasses to function effectively on-site, simple location data is insufficient. GNSS-based location technologies, represented by GPS, can have errors of several meters in typical environments, and accuracy drops significantly indoors or in complex urban areas. To provide accurate information to workers, smart glasses must determine "which equipment they are in front of and in which direction they are looking."
Visual Positioning Systems (VPS) offer a solution to this limitation. The visual sensors in smart glasses or mobile devices recognize features of the surrounding environment and compare them with pre-established spatial data to calculate location and direction with centimeter-level precision. While GNSS excels at approximate location tracking over large areas, VPS enables precise location recognition in specific spaces. The demands of the physical AI era require a comprehensive understanding of equipment, structures, work points, and pathways in three dimensions.
The statement at NVIDIA's GTC 2026 that "the factory itself is now a robotic system" reflects this context. To operate an entire factory as a single intelligent system, every space within it must be digitally understood with precision. In the past, creating accurate spatial information required expensive LiDAR scanners and skilled surveyors, making it the domain of large projects. Now, real-time spatial mapping (SLAM) can be achieved with smart glasses without special equipment. Advances in computer vision and spatial computation technology have significantly lowered this barrier.
NVIDIA's Physical AI Data Factory Blueprint provides an architecture that generates large-scale synthetic data from limited real-world data, accelerating the development of vision AI agents.
The value of technology ultimately proves itself in the field. Smart glasses represent the final point of contact where all the technology stacks of physical AI meet people. In logistics, for instance, a worker wearing smart glasses can see picking locations, product information, and quantities displayed in real-time, while vision AI automatically recognizes barcodes and integrates with the warehouse management system (WMS) instantly. This can be implemented with software alone, without additional equipment investment. In maintenance, when the smart glasses camera recognizes equipment tags or QR codes, schematics, manuals, and inspection histories are immediately displayed in the worker's view. Tasks such as generating reports can be completed using voice commands, standardizing maintenance work that previously relied on individual skill levels.
In the physical AI era, future industrial competition will hinge on how precisely the entire physical space can be digitized and how seamlessly AI can connect people, robots, and equipment. At the forefront of this connection are smart glasses that equip industrial workers with AI.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.