ETRI to unveil 9 advanced AI technologies, from disaster prediction to robot learning

The Electronics and Telecommunications Research Institute (ETRI) said Monday it will showcase nine advanced AIoT technologies — combining AI with the Internet of Things — at the 2026 AIoT International Exhibition, which opens Tuesday at COEX in Seoul for a three-day run.

AIoT refers to technology in which AI analyzes and acts on data collected from objects and sensors in real time to deliver services. Rapid advances in generative AI and large language models (LLMs) have accelerated its adoption beyond industrial settings into fields such as healthcare, transportation and energy.

ETRI will present core technologies across five areas: digital twin, robotics, future mobility, healthcare and safety, and carbon and energy management.

The flagship technology on display is an "intelligent digital federated twin" system that links multiple digital twins — virtual replicas of real-world objects or spaces — to integrate data analysis and forecast future conditions. The system is designed to assess environmental changes and the likelihood of disasters before they occur.

Also drawing attention is a large-scale on-device AI object and network management technology. Built on on-device AI — which runs AI processing directly on individual devices — the system enables multiple AI-equipped devices to share information and work together.

In robotics, ETRI will unveil Roji-Hub, a robot behavior learning dataset construction technology that systematically collects and manages motion data to help robots learn tasks. It serves as a foundational technology for physical AI, enabling robots to learn to perform a wide range of tasks in real-world environments.

ETRI will also present a positioning technology for advanced air mobility (AAM) that estimates an aircraft's location using visual data and spatial information even in environments where Global Navigation Satellite System (GNSS) signals are unavailable.

In the healthcare and safety field, ETRI will introduce an AI-powered unmanned surveillance and integrated control technology that uses cameras and LiDAR sensors to automatically detect people and vehicles.

Particularly notable is a technology that uses AI to analyze patients' vital signs in moving environments such as ambulances and assist in diagnosing cerebrovascular conditions. The system is expected to support faster decision-making by emergency medical personnel in the field.

On the carbon neutrality front, ETRI will present a product lifecycle carbon footprint management platform that tracks carbon emissions from production through disposal for products such as batteries and vehicles, alongside a technology that uses AI to analyze energy consumption and carbon emissions at data centers and industrial sites.

A smart livestock facility technology will also be on display, capable of monitoring barn environments in real time to prevent hazards such as fire.

ETRI projects that AIoT will become a core enabling technology for both "agentic AI" — systems that carry out tasks autonomously — and physical AI, which moves and acts in the real world.

"The essence of AIoT is not connecting more things, but having AI analyze data collected on-site to create new value," said Park Jun-hee, director of ETRI's ADX Convergence Research Institute. "We will work to develop technologies that solve problems in industrial settings and that people can feel in their everyday lives."