ETRI unveils six AI technologies linking physical and virtual worlds

동아사이언스· June 20, 2026

The Electronics and Telecommunications Research Institute (ETRI) has debuted six core AI-driven technologies designed to bridge the gap between physical and virtual environments at the 2026 Korea Virtual Convergence Industry Expo. These innovations focus on intelligent interaction, immersive visualization, and spatial expansion, representing a significant leap forward for the telecommunications sector's pursuit of spatial computing. By integrating AI with high-speed data processing and real-time 3D rendering, these developments provide a foundation for next-generation communication services and hyper-realistic metaverse applications.

ETRI’s showcase at the Seoul-based expo highlighted advancements in intelligent interaction through virtual AI human agents and haptic technology. The institute demonstrated an AI agent capable of responding naturally to human speech, actions, and context, moving beyond the rigid responses of previous virtual avatars. Complementing this is a new active force-feedback haptic glove that transmits precise tactile sensations and force changes when interacting with virtual objects. For the telecommunications industry, these tools signal a shift toward more intuitive and sensory-rich remote communication interfaces.

In the realm of immersive visualization, ETRI introduced technologies aimed at handling massive data loads with significantly improved efficiency. The institute unveiled an ultra-fast visualization system for 1 TB-scale physical simulation data, which utilizes AI lightweighting and multi-node parallel processing to increase frame rates from 9 fps to 20 fps. Additionally, an on-premise AI tool was presented for generating special effects (VFX) from simple user prompts. These improvements in processing speed and automated content creation are critical for telecom operators looking to deliver high-bandwidth, low-latency virtual reality and augmented reality content over next-generation networks.

The final focus area, three-dimensional spatial expansion, addresses the complexities of stereoscopic video production and broadcasting. ETRI showcased a single-camera-based free-viewpoint video generation technology, which eliminates the need for the dozens of cameras typically required for such content. Furthermore, the institute demonstrated a real-time broadcasting system that tracks multiple objects across large spaces using standard cameras to deliver multi-viewpoint stereoscopic video. Lee Jeong-ik, head of ETRI’s Hyper-realistic Metaverse Research Laboratory, emphasized that these technologies are paving the way for an era where AI and space are fully integrated, offering new commercialization and collaboration opportunities for the global telecommunications and tech industries.

Read the full story at 동아사이언스

Summary generated by RabbitReport AI from public reporting. The full article and original reporting belong to 동아사이언스.