ugo and FastLabel Launch 'ugo VLA Model Development Training Program' for Physical AI Developers
FastLabel and ugo have launched a practical training program using the 'ugo Pro R&D' humanoid robot, allowing companies to experience the end-to-end VLA model development process, from task design to real-world evaluation.
📋 Article Processing Timeline
- 📰 Published: May 27, 2026 at 10:00
- 🔍 Collected: May 31, 2026 at 22:58 (108h 58m after Published)
- 🤖 AI Analyzed: May 31, 2026 at 22:59 (1 min after Collected)
FastLabel Inc. (Headquarters: Shinjuku-ku, Tokyo; CEO: Takeshi Suzuki) and ugo Inc. (Headquarters: Chiyoda-ku, Tokyo; CEO: Ken Matsui), a developer of domestic AI robots, have announced the launch of the 'ugo VLA Model Development Training Program powered by FastLabel.' This practical program is designed to help companies, universities, and research institutions engage in Vision-Language-Action (VLA) model development from the early stages using the 'ugo Pro R&D' humanoid robot.
This training program provides an end-to-end experience, covering task design, data collection for imitation learning, data quality management, VLA model development (fine-tuning), real-world evaluation, and reporting. Even companies without existing internal expertise or development structures for physical AI can experience the development and verification process using robot arms in a short period, facilitating concrete considerations for in-house AI implementation.
[Background]
In recent years, against the backdrop of a declining birthrate, aging population, and labor shortages, improving efficiency and labor-saving in fields such as security, inspection, transport, and guidance have become significant social issues. In this context, physical AI, including VLA models that integrate vision, language, and action to flexibly control robots, is attracting attention as a next-generation technology capable of adapting to diverse on-site conditions that were difficult for conventional rule-based automation to handle.
However, when companies consider utilizing physical AI, they need an environment where they can trial the entire process from robot procurement and data collection to model development and evaluation. In reality, challenges such as 'not knowing where to start,' 'lacking internal development/verification expertise,' and 'high initial burden to launch a PoC' have acted as major barriers to early-stage adoption.
To address these issues, ugo and FastLabel have jointly developed a corporate training program that integrates domestic humanoid robots, data collection/preparation, VLA model development, evaluation, and lectures. This allows companies to move beyond theoretical discussions and grasp the potential of physical AI and the key points for their own applications through actual robot-based development and verification.
[Program Details]
The 'ugo VLA Model Development Training Program' packages the essential processes for initial VLA model verification. It provides consistent support from setup and environment construction of the 'ugo Pro R&D model' to task design, data collection, model training, deployment, evaluation, and reporting. The program is structured to accumulate foundational knowledge and practical expertise within the company through lectures and hands-on experience.
This program emphasizes ongoing support with an eye toward future in-house development and full-scale implementation. By leveraging the best practices and formats held by ugo and FastLabel, companies can aim to conduct initial development and verification processes and accumulate knowledge in as little as three months from the initial inquiry.
[Future Outlook]
ugo and FastLabel have been collaborating through research and development and sales activities in the AI robotics field. Through this training program, they aim to create an environment where more companies can start initial verification of physical AI development, contributing to the expansion of robot utilization in Japan.
This training program provides an end-to-end experience, covering task design, data collection for imitation learning, data quality management, VLA model development (fine-tuning), real-world evaluation, and reporting. Even companies without existing internal expertise or development structures for physical AI can experience the development and verification process using robot arms in a short period, facilitating concrete considerations for in-house AI implementation.
[Background]
In recent years, against the backdrop of a declining birthrate, aging population, and labor shortages, improving efficiency and labor-saving in fields such as security, inspection, transport, and guidance have become significant social issues. In this context, physical AI, including VLA models that integrate vision, language, and action to flexibly control robots, is attracting attention as a next-generation technology capable of adapting to diverse on-site conditions that were difficult for conventional rule-based automation to handle.
However, when companies consider utilizing physical AI, they need an environment where they can trial the entire process from robot procurement and data collection to model development and evaluation. In reality, challenges such as 'not knowing where to start,' 'lacking internal development/verification expertise,' and 'high initial burden to launch a PoC' have acted as major barriers to early-stage adoption.
To address these issues, ugo and FastLabel have jointly developed a corporate training program that integrates domestic humanoid robots, data collection/preparation, VLA model development, evaluation, and lectures. This allows companies to move beyond theoretical discussions and grasp the potential of physical AI and the key points for their own applications through actual robot-based development and verification.
[Program Details]
The 'ugo VLA Model Development Training Program' packages the essential processes for initial VLA model verification. It provides consistent support from setup and environment construction of the 'ugo Pro R&D model' to task design, data collection, model training, deployment, evaluation, and reporting. The program is structured to accumulate foundational knowledge and practical expertise within the company through lectures and hands-on experience.
This program emphasizes ongoing support with an eye toward future in-house development and full-scale implementation. By leveraging the best practices and formats held by ugo and FastLabel, companies can aim to conduct initial development and verification processes and accumulate knowledge in as little as three months from the initial inquiry.
[Future Outlook]
ugo and FastLabel have been collaborating through research and development and sales activities in the AI robotics field. Through this training program, they aim to create an environment where more companies can start initial verification of physical AI development, contributing to the expansion of robot utilization in Japan.
FAQ
How does this program address Japan's labor shortage?
By training experts in physical AI, it enables the automation of on-site tasks such as security and inspection, helping to mitigate the impact of a shrinking workforce.