2026: The Year of Physical AI – A Global 60 Trillion Yen Market, Japan Aims for 20 Trillion Yen with National Strategy; RX Japan Launches First-Ever Specialized 'Physical AI Expo' at Manufacturing World Tokyo

RX Japan LLC will launch the first specialized exhibition dedicated to Physical AI, titled '1st Physical AI Expo – Development and Implementation of AI × Robotics × Sensors,' at Tokyo Big Sight from Wednesday, July 1 to Friday, July 3, 2026.

AI has finally begun to move in the real world.

Humanoid robots working in factories, autonomous transport robots navigating logistics warehouses—AI is increasingly taking over tasks such as 'transporting,' 'inspecting,' and 'driving' on behalf of humans. AI, which previously evolved within digital spaces through chat and image generation, is now integrating with robots and machinery to enter the physical world. This technological domain, where AI makes decisions and operates autonomously in real space, is known as 'Physical AI.'

Behind this trend are multifaceted challenges, including severe labor shortages in logistics, manufacturing, construction, and infrastructure, as well as the need to pass on skilled expertise and replace hazardous tasks. Physical AI is being implemented not only in humanoid robots but also in robotic arms, drones, and autonomous vehicles, drawing attention as the next-generation industrial infrastructure.

The Ministry of Economy, Trade and Industry (METI) estimates that the AI robotics market, including humanoid robots, will grow to approximately 60 trillion yen globally by 2040*. Japan has also set a national strategy aiming to capture over 30% of the global market and achieve a 20-trillion-yen domestic market, marking 2026 as a pivotal year—potentially the 'Year One of Physical AI.'

In response to this momentum, RX Japan has decided to launch the specialized 'Physical AI Expo.'

The exhibition will feature live demonstrations such as humanoid robots performing tasks, autonomous transport robots in motion, real-time spatial recognition by AI, and quadruped robots—allowing visitors to see and experience the 'workplace of the future' firsthand.

This inaugural event will showcase a future where AI not only 'thinks' but also 'acts' in the real world, enabling humans and robots to work together.

*Reference: METI 'Reference Materials for AI Robotics Study Group' (October 2025), 'AI Robotics Strategy' (March 26, 2026)

What is Physical AI?

Physical AI refers to AI-powered robots capable of directly interacting with physical environments and performing tasks flexibly and adaptively, much like humans. Unlike conventional AI technologies that have excelled in cyberspace, Physical AI systems are expected to create new value by solving real-world challenges.

(Source: Research and Development Strategy Center (CRDS), Japan Science and Technology Agency (JST), Workshop Report 'Physical AI Systems')

■ Differences Between Generative AI and Physical AI

If Generative AI is the 'thinking AI,' then Physical AI is the 'working AI in the real world.'

Generative AI

Physical AI

Field of Operation

Digital Space

(PCs, smartphones, cloud, etc.)

Real World

(factories, logistics, on-site operations, etc.)

Primary Role

Generate text, images, code /

Think and respond

Move, transport, perform tasks /

Make decisions and act

Representative Examples

Efficiency in intellectual work

Addressing labor shortages, automating on-site operations

Social Impact

Chat AI, image generation AI, etc.

Humanoid robots, autonomous transport and inspection robots, etc.

Exhibited Products Introduction (excerpt)

Company: Prox Industries Inc.

Robot technology that moves like a human hand

Live demonstrations of UR3 dual-arm and Franka × Inspire Hand (5-fingered) will be showcased. Multiple demos developed specifically for the exhibition will illustrate practical applications of Physical AI.

Company: DIC Corporation

Achieving delicate and precise grasping, just like a human hand

A robot finger that spontaneously recognizes workpieces without cameras or sensors, capable of grasping any object. Not only can its shape and sensors be freely customized, but through simple teaching programming, it autonomously recognizes and grasps desired objects.

Company: Ryosan Hyoyo Co., Ltd.

Supporting the first step toward Physical AI adoption – Desktop Robot

A robot made with a 3D printer, combined with NVIDIA inference environment (hardware for AI inference) and GR00T N1.5 (robot base model). Visitors can experience Physical AI as the robot autonomously makes decisions and acts in response to environmental changes.

Company: PonoSHIP Inc.

Sim-to-Real: Achieving Autonomous Mobility through Physical AI

A solution enabling autonomous patrols and data collection in factories, security, and equipment maintenance. By rapidly training in simulation, it safely and cost-effectively implements advanced functions such as obstacle avoidance, stair climbing, and anomaly detection, contributing to labor shortage mitigation and improved safety.

Company: Kyocera Corporation

High-precision automatic recognition of mixed-load cases with different shapes using proprietary 3D × AI vision system

Even mixed cardboard boxes that are difficult to recognize can be accurately located using Kyocera’s unique AI recognition technology. Regardless of changes in box size, color, shape, or stacking method, container and bag loading/unloading and pallet reconfiguration can be performed without program modifications.

Company: Airion Inc.

Fully automating container movement between storage racks

A University of Tokyo-originated AI venture focused on 'intelligent' machinery in the manufacturing FA (Factory Automation) sector. At the exhibition, a live demo will showcase the latest humanoid robot autonomously transporting items to locations specified by attendees.

■ Press Inquiry (Pre-registration Required)

Press registration link:

https://www.manufacturing-world.jp/tokyo/ja-jp/media/press-registration.html

―― Please feel free to contact us ――

• We welcome advance consultation on interview topics and targets

• Press interviews on the day before the event (June 30) may be accommodated upon schedule coordination

• We will do our best to accommodate press inquiries on event days

Press opportunities at the exhibition include:

• Coverage of latest manufacturing trends: Physical AI, AI robotics, automation, labor-saving

• Seminar reporting on key themes (featuring NVIDIA, FANUC, Yaskawa Electric, etc.)

• Interviews with organizers, exhibitor booths, attendees, and on-site venue photography

Event Overview

First-time launch within Japan’s largest manufacturing exhibition, 'Manufacturing World'!

Exhibition Name: 38th Manufacturing World [Tokyo]

1st Physical AI Expo – Development and Implementation of AI × Robotics × Sensors

Dates: July 1 (Wed) – 3 (Fri), 2026, 10:00 – 17:00

Venue: Tokyo Big Sight (entire facility)

Organizer: RX Japan LLC

Exhibitors: ~2,000 companies (planned)

Visitors: ~70,000 (estimated)

(Note: Figures for the entire Manufacturing World event)

Admission: Free (prior visitor registration required)

Official Website: https://www.manufacturing-world.jp/tokyo/ja-jp.html