NVIDIA is accelerating the implementation of real-time intelligent systems by unveiling a next-generation physical AI platform for industrial sites.

NVIDIA announced on the 17th (local time) that it officially launched its industrial edge AI platform, 'IGX Thor,' at 'GTC 2026' held in San Jose, California.

As industries across the board move beyond simple automation into the stage of physical AI equipped with autonomy and intelligence, there is a growing need for high-performance computing systems capable of making immediate decisions and performing safe control on-site.

In response to these demands, IGX Thor integrates high-speed sensor data processing capabilities, enterprise-grade reliability, and functional safety into a single platform.

IGX Thor is designed to analyze large-scale data collected from various sensors in real time and immediately reflect the results of AI inference. This enables stable AI operation even in safety-critical environments, such as autonomous driving equipment, industrial robots, and medical devices.

IGX Thor is expanding its scope of application to include construction, manufacturing, logistics, healthcare, life sciences, as well as space exploration.

Heavy equipment manufacturer Caterpillar is introducing an in-vehicle AI assistant based on the IGX Thor to support worker decision-making and simultaneously improve on-site safety and productivity.

In the railway sector, Hitachi Rail is building predictive maintenance and autonomous inspection systems using IGX Thor. Through this, the stability of the railway network is being strengthened and operational efficiency is being improved.

In the field of logistics automation, KION Group has enhanced the perception capabilities of autonomous robots by combining IGX Thor and NVIDIA's safety workflow.

The safety of the work environment is being enhanced by precisely controlling robot movements using cameras installed in the infrastructure and virtual safety zones.

Robotics companies Agility and Hexagon Robotics are enhancing real-time AI inference and multimodal sensor fusion technology for humanoid robots by applying IGX Thor.

This has enabled stable robot operation even in complex environments.

The use of IGX Tor is also expanding in the medical field.

Johnson & Johnson is operating a digital surgery platform that applies real-time AI inference to the operating room, and Karl Storz is developing next-generation endoscopes and diagnostic imaging equipment based on the IGX Thor.

It is assessed that ultra-low latency AI processing has become possible in surgical environments where precision and safety are required.

Planet Labs is utilizing IGX Tor to efficiently analyze vast amounts of satellite data and convert it into practical information.

The European Organization for Nuclear Research (CERN) has also adopted IGX Tor to process large-scale scientific data at high speeds and is operating advanced AI models.

NVIDIA is building a collaborative ecosystem with semiconductor, sensor, and system manufacturers to expand IGX Tor.

Major semiconductor companies are integrating sensor and motor control technologies with NVIDIA's simulation and sensor integration framework, and industrial system manufacturers are developing IGX Thor-based products tailored to various environments.

NVIDIA stated, “Physical AI is the core that simultaneously boosts safety and productivity in industrial sites.”Describing it as “shim technology,” he stated, “IGX Thor will become a new standard in all industrial environments requiring real-time AI.”