The University of Tokyo, NTT, Inc., and NEC Corporation have successfully integrated three innovative technologies on a 6G/IOWN platform to enhance the deployment of AI agents aimed at bolstering societal safety and security. The collaboration, spearheaded by The University of Tokyo’s Social Cooperation Program, leverages streaming semantic communication, AI-oriented media control, and In-Network Computing (INC) architecture technologies.
This initiative addresses critical challenges in data transmission and computational processing, as AI agents increasingly rely on high-volume data from sensors and devices. The three parties conducted a trial to verify the efficiency of these technologies, confirming their potential to reduce latency and computational load while maintaining AI inference accuracy.
The technologies will be showcased at the Mobile World Congress 2026 in the Japan Pavilion, where the research findings will be presented to an international audience. This collaboration underscores the importance of international cooperation and standardization in advancing telecommunications infrastructure.
Streaming semantic communication technology, developed by The University of Tokyo, focuses on transmitting only semantic differences, significantly reducing the use of wireless communication resources. NEC's AI-oriented media control technology assigns data identifiers to AI agents, selectively feeding key sensor data to reduce computational demands. NTT's In-Network Computing architecture enhances AI processing by distributing small, specialized AI models across the network.
The trial involved using a video dataset to assess the impact of these technologies on end-to-end latency, a critical factor for real-time AI applications. The results demonstrated a stable reduction in latency and computational load without compromising AI accuracy, making these technologies suitable for scenarios requiring real-time performance, such as augmented reality applications.
The University of Tokyo, NEC, and NTT each played distinct roles in the development of these technologies. The University of Tokyo focused on bandwidth constraints, NEC tackled computational load challenges, and NTT addressed power consumption and scalability issues. Their combined efforts have resulted in a promising solution for the widespread use of AI agents in future 6G networks.
Looking ahead, the three organizations plan to accelerate research and development to implement these findings socially, aiming to realize AI agents and ICT infrastructure that enhance safety and security. The initiative highlights the potential of AI agents to autonomously perceive and respond to environmental changes, providing a proactive approach to disaster response, cyberattacks, and other critical situations.