Active Club’s Newsletter brings you the latest news and trends in manufacturing, robotics, and cutting-edge technology.
🗞️ This Week in Robotics
Gemini Robotics brings AI into the physical world
Google DeepMind has introduced Gemini Robotics, an AI model designed for robotics. This model integrates vision and language to create a general-purpose AI system that can adapt to various robots. While still in the research phase, it aims to help robots perform physical tasks autonomously and efficiently.
AgiBot GO-1: The Evolution of Generalist Embodied Foundation Model from VLA to ViLLA
AgiBot has unveiled Genie Operator-1 (GO-1), a physical AI foundation model that utilizes a method called Vision-Language-Latent-Action (ViLLA) to understand the physical world, plan actions, and execute tasks.
GO-1 is part of the AGIBOT World Colosseo platform, which provides open-source datasets, tools, and models to democratize scalable and general-purpose physical AI systems.
📈 Market & Business
Japan’s service robot market projected to triple in five years
Japan's service robot market is projected to triple to ¥400 billion ($2.7 billion) by 2030.
As businesses face an aging population and labor shortages, service robot adoption is accelerating. In the restaurant chain Skylark, around 3,000 cat-ear robots deliver food, while 71-year-old employees work alongside robots for efficiency, demonstrating real-world impact.
🧰 Resource & Tools
Manus, a New AI Agent From China is Going Viral—And Raising Big Questions
A new Chinese AI agent, Manus, is gaining attention for its ability to analyze, plan, and execute tasks autonomously beyond traditional AI assistants.
Using a multi-agent system, Manus is being applied to resume screening, real estate research, and stock analysis. Industry leaders, including Hugging Face, have praised it as a "promising AI agent."
New tools for building agents
OpenAI has released new tools for AI agent development, including:
Responses API
Built-in tools for web and file search
Agents SDK for orchestrating core agent logic
Introducing Gemma 3: The most capable model you can run on a single GPU or TPU
Google AI has launched Gemma 3, available in 9B and 27B sizes, supporting up to 1 million tokens for long-context inference. Designed as an open model, it runs efficiently even on limited hardware resources.
🌈 Others
The future of robotics: 7 predictions for 2025 and beyond
This report explores seven key predictions for the future of robotics across manufacturing, logistics, healthcare, homes, environmental conservation, and education.
Key trends include:
Cobots (collaborative robots) working alongside humans
AI-powered robots making autonomous decisions
Affordable robotics for SMEs to improve efficiency and safety
While robotics adoption is accelerating, the report also highlights ethical challenges, cost concerns, and privacy issues, emphasizing the need for societal acceptance and infrastructure development.
Stay tuned for the next issue!
✍️ @nszknao_jp