FastLabel Inc. (Headquarters: Shinjuku-Ku, Tokyo; President and CEO: Takeshi Suzuki; hereinafter “FastLabel”), which supports the development of robot foundation models and domain-specific Vision-Language-Action (VLA) models, today announced the establishment of the Frontier AI Fundamental Technology Development Group under its Robotics AI Business Division. This group has been created to further strengthen the company's technology development capabilities in the AI robotics field. The group is not only conducting research and development but also advancing AI robotics technologies through the integration of foundational technologies and data, while promoting the entire process from research to real-world deployment to accelerate the realization of Physical AI in society. Background to the Establishment of the New Group Since its founding, FastLabel has supported AI robotics development for a wide range of organizations, including leading Japanese enterprises and research institutions, through a data-centric approach to AI development. The company has provided end-to-end support across all stages of the AI development lifecycle, including data collection, data annotation and processing, model development, and evaluation. In recent years, the AI robotics field has witnessed rapid advancements in model architectures, particularly in the United States and China. As these technologies continue to evolve, the types of data required—and the methods used to collect them—are changing at an unprecedented pace. At the same time, Japan has been accelerating efforts to strengthen its global competitiveness in AI robotics. Under government leadership, initiatives such as the AI Robotics Strategy are promoting collaboration between the public and private sectors to drive early real-world deployment of AI robotics technologies and establish Japan's position as a global leader in the field. Against this backdrop, FastLabel believes that realizing truly valuable applications of Physical AI requires more than simply leveraging state-of-the-art models. Success depends on a comprehensive approach that integrates diverse elements, including robotics control technologies, a deep understanding of model capabilities and limitations, and sophisticated methods for collecting, managing, and utilizing high-quality data. The company views the ability to orchestrate these components holistically as a critical factor in achieving successful real-world implementation. The establishment of the Frontier AI Fundamental Technology Development Group is intended to enable FastLabel to evolve beyond a conventional research and development (R&D) organization into an organization that creates value through AI robotics development based on foundational technologies and data, with a view toward the social implementation of Physical AI. Leadership of the Robotics AI Business Division Division Head: Hiroki Fujiwara Executive Officer & Chief Operating Officer (COO) Department Head: Kenichiro Tsukamoto General Manager, Business Development Department, Yudairo Miyake General Manager, Data Solutions Department About the Newly Established Frontier AI Fundamental Technology Development Group Organization Name: Frontier AI Fundamental Technology Development Group Data Solutions Department, Robotics AI Business Division Launch Date: May 1, 2026 Group Head: Nao Yamada Data Solutions Department, Robotics AI Business Division Examples of Ongoing Projects The Frontier AI Fundamental Technology Development Group conducts research, evaluation, and validation of technologies spanning from foundational to cutting-edge AI robotics technologies to address customer challenges. Some of our initiatives are introduced in the following tech blog. ■Understanding π*0.6: A Reinforcement Learning Method for Robotics AI This article explains the overall framework of RECAP, which is used in Physical Intelligence's reinforcement learning method π0.6, and explores in depth how the value function and policy updates are designed. ■What Is QDD? The Actuator Technology Gaining Attention in Robotics This article explains, from the ground up, the mechanisms of motors and gear reducers and the importance of backdrivability, all of which are essential for enabling AI to control robots. It also organizes the physical characteristics and advantages of Quasi-Direct Drive (QDD), which is becoming the mainstream approach in next-generation robotics. ■Developing a VLA Model with ugo Pro This article explains an overview of VLA model development using the imitation learning kit for the Japanese-made robot ugo Pro. It also explains ugo Pro, tips for data collection, and methods for dataset construction. Recruiting Physical AI Engineers With the launch of the Frontier AI Fundamental Technology Development Group , FastLabel is actively seeking talented engineers who are passionate about accelerating the real-world deployment of AI robotics technologies. Key Responsibilities Research, evaluate, and validate state-of-the-art robot foundation models, learning strategies, and data collection methodologies Train and evaluate robot foundation models and develop robotic control systems powered by these models Design and build advanced data collection frameworks and systems Solve customer challenges through the deployment of AI robotics solutions in real-world environments We are looking for professionals who are committed to technology and customers and who aspire to solve real-world social challenges together with us. Career Opportunities: https://herp.careers/v1/fastlabel/dHh35DUK1nXO About FastLabel FastLabel, Inc. is committed to building data infrastructure that enables data-centric AI development. The company provides end-to-end support across the AI development lifecycle, including data collection and generation, annotation, model development, and DataOps implementation. In recent years, FastLabel has expanded its focus to the field of physical AI, including robotics, and has been developing data pipelines that support the creation of robot foundation models and Vision-Language-Action (VLA) models. Company Overview Company Name: FastLabel Inc. Headquarters: Shinjuku Sumitomo Building 24F, 2-6-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo, Japan Representative: Takeshi Suzuki, President and CEO Business: Provider of professional services and products supporting Data-centric AI development Website: https://fastlabel.ai/service/robotics