
Following media coverage of the government's growth strategy, attention to "Physical AI" is increasing. Physical AI is a technological domain that connects AI with robots, equipment, and sensors to enable real-world operations and control. This article organizes the direction of Physical AI that the government is pursuing from a DX perspective, explains the key points necessary for implementation, and discusses areas where Haudi can contribute.
News Overview
In June 2026, the Nikkei reported that the government's growth strategy includes a plan to invest ¥10.5 trillion in Physical AI through public-private collaboration by fiscal year 2040. This "¥10.5 trillion" is not the amount the government will spend as subsidies, but rather the combined investment outlook for both public and private sectors, including corporate R&D, capital investment, and implementation, along with government investment.
The article explains Physical AI as technology that uses AI to autonomously operate robots and other devices. The scale of investment suggests that the government positions Physical AI as one of the priority areas in its growth strategy.
〈Reference Article〉Nikkei "17 Strategic Fields, ¥10.5 Trillion for Physical AI - Full Details of Public-Private Investment Revealed"
What the Government is Pursuing
The government aims to expand AI utilization beyond digital space into industrial Sites and social infrastructure.
This is driven by challenges such as labor shortages, skilled worker shortages, and productivity improvement. In fields such as manufacturing, logistics, construction, and infrastructure management, AI and robots are expected to support or replace some of the work traditionally performed by humans, enhancing safety and efficiency.
Government documents also highlight key issues such as social implementation of AI robotics, utilization of Sites data, improvement of AI models, and ensuring safety and security.
〈Reference Materials〉
Cabinet Secretariat "Public-Private Investment Roadmap for 'Key Products and Technologies' in 17 Strategic Fields"
Cabinet Office "AI and Semiconductor Working Group (2nd Meeting) Distributed Materials"
•Material 4-1 "AI Robotics Strategy Overview"
•Material 4-2 "AI Robotics Strategy"
•Material 4-3 "AI Robotics Implementation Roadmap by Field"
What the government is pursuing is not just R&D in AI robotics. It can be understood as an initiative to make AI actually function in industry and social infrastructure by utilizing data obtained from Sites and connecting it to decision-making and control.
What is Physical AI?
Physical AI is a technological domain where AI recognizes and makes decisions based on Sites data, connecting to the operation of equipment, devices, and robots. While generative AI handles digital information such as text, images, and audio, Physical AI deals with real-world information and actions.
For example, by acquiring equipment status through sensors and understanding Sites conditions through cameras, AI can determine signs of change or the presence of anomalies. Based on these judgments, robots move, transport systems change routes, or air conditioning temperature and airflow settings are adjusted.
The key characteristic of Physical AI is that AI not only analyzes and makes decisions about information, but also connects those results to Sites movements and control.

〈Related Information〉
For those who want to learn more about Physical AI, please also see Haudi's "Introduction to Physical AI" series. It explains the technical background, application fields, and challenges for social implementation in five parts.
Part 1: What is Physical AI - Overview and Background of Its Emergence
Part 2: Sensing and Edge AI - The "Eyes and Brain" Supporting Physical AI
Part 3: Industrial Applications of Physical AI - Robotics and Autonomous Systems
Part 4: Physical AI in Smart Cities - Infrastructure that "Senses and Moves" the City
Part 5: The Future and Challenges of Physical AI - Ethics, Safety, and Social Acceptance
What is Needed for Sites Implementation of Physical AI
Sites implementation of Physical AI is an extension of the data utilization that corporate DX has been aiming for. It cannot be achieved simply by introducing AI models or robots.
First, it is necessary to properly understand the state of Sites. Which equipment is operating and how? Where is the workload concentrated? What data is being acquired and what data is lacking? How will existing equipment be connected to new systems? Such organization is the starting point.
On that basis, the following elements need to be combined:
- Sensors and IoT devices to acquire Sites data
- Connection with existing equipment and devices
- Mechanisms to securely transfer data
- Data processing in the cloud or at the edge
- AI-based decision-making and prediction
- Control mechanisms to operate equipment and devices
- Operational design considering security and safety
Physical AI is not something that can be completed with AI alone. It is a comprehensive initiative to collect Sites data, organize it into a usable form, and connect it to equipment and device operations as needed.
Areas Where Haudi Can Contribute
To utilize Physical AI, it is important to first establish a foundation for utilizing data in business operations and equipment Sites before introducing AI systems or robots. This is an essential initiative not only for Physical AI but also for embedding corporate DX in Sites.
Haudi supports the creation of such data utilization foundations by leveraging technologies in IoT, AI, and information security. From consultation at the concept stage to IoT devices, data transfer, cloud integration, control, and automation, we propose solutions tailored to Sites conditions. Services for acquiring and transferring data from existing equipment and devices that are difficult to connect to networks, as well as services for decision-making and control aimed at building and equipment control and energy optimization, are part of these initiatives.
Physical AI does not only mean introducing advanced AI robots in one leap. Initiatives such as acquiring and visualizing Sites data, connecting with equipment, and small-scale automation are also important steps toward future Physical AI implementation.
Haudi supports the first steps toward Physical AI implementation through the creation of data utilization mechanisms tailored to business operations and equipment Sites.
Summary
Physical AI is a keyword for an era when AI goes beyond digital space to support real-world Sites. As it is positioned as a priority field in the government's growth strategy, it becomes important for companies to utilize Sites data and connect it to business improvement and automation.
Realizing Physical AI does not only mean introducing advanced AI robots in one leap. Collecting Sites data, organizing it into a usable form, and connecting it to necessary decisions and control - this accumulation embeds DX in Sites and leads to future Physical AI implementation.