FastLabel Strengthens Technology Development System to Address Evolving AI Robotics Domain ~ Establishes Advanced AI Foundation Technology Development Organization to Accelerate Social Implementation of Physical AI Through Development of Foundational Technologies and Data ~
FastLabel has established the "Advanced AI Foundation Technology Development Group" to strengthen its technology development in the AI robotics domain. This aims to support the development of robot foundation models and VLA models, accelerating the social implementation of physical AI.
📋 Article Processing Timeline
- 📰 Published: May 11, 2026 at 18:00
- 🔍 Collected: May 11, 2026 at 09:31
- 🤖 AI Analyzed: May 11, 2026 at 10:51 (1h 19m after Collected)
FastLabel Inc. (Headquarters: Shinjuku-ku, Tokyo; Representative Director and President: Takeshi Suzuki; hereinafter "FastLabel"), which supports the development of robot foundation models and domain-specific VLA (Vision-Language-Action) models, announces the establishment of the "Advanced AI Foundation Technology Development Group" under its Robotics AI Business Division. This move aims to further strengthen its technology development system in the AI robotics domain.
This organization will not only focus on research and development but will also engage in AI robotics technology development from the perspective of combining foundational technologies with data. By promoting an integrated approach from research to social implementation, it aims to achieve the social implementation of physical AI.
[Background of the New Establishment]
FastLabel has been supporting the AI robotics development of many companies, including major domestic corporations and research institutions, through each development phase of data collection, data processing, model development, and evaluation, centered on data-centric AI development.
In recent years, in the AI robotics domain, model architectures are evolving daily, mainly in the United States and China, and the required data and data collection methods are rapidly changing. On the other hand, in Japan, to avoid falling behind globally, the government has formulated an "AI Robotics Strategy," and public and private sectors are collaborating to achieve social implementation proactively, aiming to establish AI robotics technology and global competitiveness.
As mentioned above, in the rapidly evolving AI robotics domain, to achieve truly valuable social implementation of physical AI, it is essential to not only utilize the most advanced models but also to comprehensively integrate various elements such as robot control technology, understanding of model characteristics, and high-quality data collection and management methods. This is considered the key to success.
The establishment of this "Advanced AI Foundation Technology Development Group" aims for FastLabel to evolve into an organization responsible for creating value that looks ahead to the social implementation of physical AI, by driving AI robotics development based on foundational technologies and data, rather than merely remaining a research and development (R&D) organization.
[About the Robotics AI Business Division]
Division Head: Hiroki Fujiwara, Executive Officer COO
Manager: Kenichiro Tsukamoto, Business Development Dept., Robotics AI Business Division
Manager: Yudai Miyake, Data Solutions Dept., Robotics AI Business Division
[About the Advanced AI Foundation Technology Development Group (Newly Established)]
Name: Advanced AI Foundation Technology Development Group, Data Solutions Dept., Robotics AI Business Division
Date of Establishment: May 1, 2026
Group Leader: Naoshi Yamada, Data Solutions Dept., Robotics AI Business Division
[Examples of Initiatives]
The "Advanced AI Foundation Technology Development Group" conducts research and verification of foundational to advanced technologies in the AI robotics domain to solve technical and customer challenges. Some of these initiatives are published in the following tech blogs:
"Understanding Reinforcement Learning Method π*0.6 for Robotics AI"
This article explains the overall picture of RECAP used in the reinforcement learning method π*0.6 published by Physical Intelligence Inc., and delves into how value functions and policy updates are designed.
https://fastlabel.hatenablog.com/entry/20260508/1778226634
"What is the Actuator 'QDD' attracting attention in Robots?"
This explains, from a bottom-up perspective, the mechanisms of motors and reducers indispensable for freely controlling robots with AI, and the importance of "backdrivability." It organizes the physical characteristics and advantages of Quasi-Direct Drive (QDD), which is becoming mainstream in next-generation robotics.
https://fastlabel.hatenablog.com/entry/20260508/1778234983
"Developing VLA Models with ugo Pro"
This outlines VLA model development using the imitation learning kit for the domestic robot ugo Pro. This blog explains ugo Pro, tips for data collection, and methods for dataset construction.
https://fastlabel.hatenablog.com/entry/20260508/1778230040
[Recruiting Physical AI Engineers]
With the establishment of the "Advanced AI Foundation Technology Development Group," we are actively recruiting engineers to promote the social implementation of AI robotics.
Main roles:
・Research and verification of cutting-edge robot foundation models, learning strategies, and data collection methods
・Learning and evaluation of robot foundation models and realization of robot control using models
・Cutting-edge data collection
This organization will not only focus on research and development but will also engage in AI robotics technology development from the perspective of combining foundational technologies with data. By promoting an integrated approach from research to social implementation, it aims to achieve the social implementation of physical AI.
[Background of the New Establishment]
FastLabel has been supporting the AI robotics development of many companies, including major domestic corporations and research institutions, through each development phase of data collection, data processing, model development, and evaluation, centered on data-centric AI development.
In recent years, in the AI robotics domain, model architectures are evolving daily, mainly in the United States and China, and the required data and data collection methods are rapidly changing. On the other hand, in Japan, to avoid falling behind globally, the government has formulated an "AI Robotics Strategy," and public and private sectors are collaborating to achieve social implementation proactively, aiming to establish AI robotics technology and global competitiveness.
As mentioned above, in the rapidly evolving AI robotics domain, to achieve truly valuable social implementation of physical AI, it is essential to not only utilize the most advanced models but also to comprehensively integrate various elements such as robot control technology, understanding of model characteristics, and high-quality data collection and management methods. This is considered the key to success.
The establishment of this "Advanced AI Foundation Technology Development Group" aims for FastLabel to evolve into an organization responsible for creating value that looks ahead to the social implementation of physical AI, by driving AI robotics development based on foundational technologies and data, rather than merely remaining a research and development (R&D) organization.
[About the Robotics AI Business Division]
Division Head: Hiroki Fujiwara, Executive Officer COO
Manager: Kenichiro Tsukamoto, Business Development Dept., Robotics AI Business Division
Manager: Yudai Miyake, Data Solutions Dept., Robotics AI Business Division
[About the Advanced AI Foundation Technology Development Group (Newly Established)]
Name: Advanced AI Foundation Technology Development Group, Data Solutions Dept., Robotics AI Business Division
Date of Establishment: May 1, 2026
Group Leader: Naoshi Yamada, Data Solutions Dept., Robotics AI Business Division
[Examples of Initiatives]
The "Advanced AI Foundation Technology Development Group" conducts research and verification of foundational to advanced technologies in the AI robotics domain to solve technical and customer challenges. Some of these initiatives are published in the following tech blogs:
"Understanding Reinforcement Learning Method π*0.6 for Robotics AI"
This article explains the overall picture of RECAP used in the reinforcement learning method π*0.6 published by Physical Intelligence Inc., and delves into how value functions and policy updates are designed.
https://fastlabel.hatenablog.com/entry/20260508/1778226634
"What is the Actuator 'QDD' attracting attention in Robots?"
This explains, from a bottom-up perspective, the mechanisms of motors and reducers indispensable for freely controlling robots with AI, and the importance of "backdrivability." It organizes the physical characteristics and advantages of Quasi-Direct Drive (QDD), which is becoming mainstream in next-generation robotics.
https://fastlabel.hatenablog.com/entry/20260508/1778234983
"Developing VLA Models with ugo Pro"
This outlines VLA model development using the imitation learning kit for the domestic robot ugo Pro. This blog explains ugo Pro, tips for data collection, and methods for dataset construction.
https://fastlabel.hatenablog.com/entry/20260508/1778230040
[Recruiting Physical AI Engineers]
With the establishment of the "Advanced AI Foundation Technology Development Group," we are actively recruiting engineers to promote the social implementation of AI robotics.
Main roles:
・Research and verification of cutting-edge robot foundation models, learning strategies, and data collection methods
・Learning and evaluation of robot foundation models and realization of robot control using models
・Cutting-edge data collection