Corporate News Report
Overview
KDDI Research, Inc., a subsidiary of KDDI Corporation, has confirmed its participation in a multi‑industry initiative to develop a domestically produced multimodal foundation model for artificial intelligence. The project, spearheaded by Noetra Corp., brings together a diverse coalition of corporate and research partners—including Sony Group, SoftBank, NEC, and Honda—as well as the National Institute of Advanced Industrial Science and Technology (AIST) and Preferred Networks. The objective is to create AI models capable of simultaneously processing text, images, video, and audio, thereby enabling the deployment of intelligent systems in physical environments such as robotics, autonomous vehicles, and smart manufacturing.
Technical Contributions
KDDI Research will contribute its deep expertise in telecommunications infrastructure and data management to the creation of secure, high‑performance AI computing platforms. Leveraging NVIDIA’s state‑of‑the‑art graphics processing units (GPUs), the consortium will establish a robust computational foundation with construction slated to commence in early 2027. The development roadmap outlines the delivery of a reasoning‑focused foundation model by the end of 2026, an omni‑modal model by 2028, and a fully real‑world native AI platform by 2030.
Cross‑Sector Implications
Telecommunications and AI The integration of telecom infrastructure with AI workloads highlights an evolving convergence between network engineering and machine learning. High‑throughput, low‑latency connectivity is essential for real‑time inference in robotics and automated systems, positioning KDDI Research to become a pivotal enabler of edge AI solutions.
Manufacturing and Automation The participation of automotive and electronics giants such as Honda and Sony signals a strategic push toward intelligent automation in production lines. Multimodal AI models can enhance quality control, predictive maintenance, and supply‑chain optimization, thereby reducing operational costs and increasing resilience.
Research and Development Ecosystem Collaboration with academic and research institutions like AIST and Preferred Networks injects cutting‑edge scientific expertise into the project, fostering a virtuous cycle of innovation and talent development. This partnership model exemplifies the “innovation cluster” approach that has proven effective in technology hubs worldwide.
Economic and Strategic Context
The initiative reflects a broader trend in Japan’s industrial strategy aimed at strengthening national competitiveness in artificial intelligence. By prioritising data security, industrial applicability, and societal impact, the coalition aligns with governmental policies that emphasize ethical AI deployment and digital sovereignty.
From an economic perspective, the development of a domestic multimodal foundation model can reduce reliance on foreign AI platforms, preserve intellectual property, and create high‑skill jobs. Moreover, the project positions Japan as a leader in AI hardware‑software integration, potentially attracting foreign investment and enhancing export capabilities in AI‑enabled products.
Future Outlook
KDDI Research’s involvement underscores its commitment to leveraging telecommunications assets to accelerate the adoption of multimodal AI across Japan’s industrial landscape. The projected timelines—reasoning‑focused model by 2026, omni‑modal model by 2028, and a fully integrated AI platform by 2030—suggest a rapid maturation of capabilities that could transform service delivery in sectors ranging from healthcare to logistics.
By aligning its strategic priorities with the needs of a connected, intelligent economy, KDDI Research and its partners are poised to deliver technology that not only advances corporate efficiency but also contributes to broader societal benefits.
