Japan has long been a leader in industrial robotics, and now a new wave of startups is adding a brain to all the hardware. The field is called physical AI, meaning AI that lets robots, vehicles and industrial systems perceive and act in the real world.
The region has installed 44,500 industrial robots in 2024, according to the International Federation of Robotics, making it the world’s second-largest market behind China. That installed base is the foundation the startups below are building on, adding the software that lets machines cope with less predictable work.
The strongest companies are those pairing deep software engineering with credible pipelines into factories and fulfilment centres.
Japan Built The Robots And Now Its Startups Are Teaching Them To Think
The secret weapon is Japan’s uniquely concentrated industrial network. Because traditional robot makers, precision motor suppliers, sensor companies, chip houses and factory integrators are jammed together in one collaborative web, the transition to physical AI happens fast. Brilliant algorithms mean little until they survive contact with physical hardware, and this domestic network gives startups the ultimate testing ground.
The state is throwing its weight behind the movement too. In July 2026, SoftBank pulled together 44 enterprises including Sony, NEC and Honda to form the Noetra consortium. The group is tasked with building a homegrown AI foundation model, with initial government backing reported at approximately ¥387.3 billion.
The roadmap sets its sights on autonomous manufacturing, smart mobility, logistics, medicine and public works. It forms part of a wider Japanese tech startup movement that goes beyond just traditional hardware.
More from Tech
- From Pit Lanes To Cargo Planes: The Logistics Behind Moving Formula 1 Around The World
- AI Can Now Build A Free Adobe Rival – So Why Can’t It Open Files?
- How Modern Rendering Solutions Fit Into Creative Pipelines
- Women In Tech Week: AI Has Become A Career Skill – So Why Are Fewer Women Using It?
- Same Car, Different Driver: The Tech Behind F1’s Custom Set-Ups
- From Web3 Hype To Commercial Advantage: How The Next Phase Of Emerging Technology Will Be Won
- Who Controls The Thermal Skies? The Security And Privacy Questions Behind SatVu’s New Satellite
- The Drone Technology Helping Farmers Cover More Ground In Less Time
The Top Physical AI And Industrial Robotics Startups In Japan
Each of these eight startups takes a different route to getting AI out of the cloud and into the physical world:
Mujin
Founded in Tokyo in 2011, Mujin specialises in robot intelligence through MujinOS. The platform blends 3D perception, intelligent path planning and real-time virtual simulations to automate variable workflows on factory floors and distribution lines.
As opposed to building robots from scratch, it makes existing robot arms and equipment work in messy, changing environments, handling objects of inconsistent shapes and placements.
The company announced a $233 million Series D first close in December 2025, which it says will support global expansion. Because it sells a universal software layer capable of running across various industrial robots instead of locking customers into a proprietary machine, it has become one of Japan’s most prominent physical AI startups.
Telexistence
Tokyo-based Telexistence, founded in 2017, builds AI-enabled robots that combine autonomous operation with remote human support.
Its TX SCARA robot has been used in FamilyMart stores for beverage replenishment, with remote operators stepping in for tasks that image recognition can’t reliably complete.
In July 2023, the company reeled in about ¥23 billion in Series B financing from investors including SoftBank Group, Foxconn and Globis Capital Partners. Its technology has since caught the attention of Physical Intelligence, which is working with the company on general-purpose robot foundation models using retail tele-operation data.
Rapyuta Robotics
Born out of ETH Zurich research and now headquartered in Tokyo, Rapyuta Robotics builds autonomous mobile vehicles, cloud-connected robotics software and automated storage-and-retrieval systems designed specifically for modern warehouses.
Its products include the rapyuta.io cloud robotics platform, the Rapyuta ASRS and the Rapyuta PA-AMR picking-assist robot, which link robot fleets to warehouse-management systems, inventory data and human operators.
In May 2025, the company said its ASRS had begun full-scale operation at an Alps Logistics warehouse. By tying the machinery into site operations, the platform moves away from treating robots as disconnected silos on the warehouse floor.
TIER IV
Founded in 2015 and based in Tokyo, TIER IV is the company behind Autoware, an open-source autonomous-driving software stack used in autonomous shuttles, delivery systems and industrial machinery.
The company’s strategy is to provide the software foundation for autonomous vehicles rather than build every vehicle itself. Its Series B total reached ¥20.7 billion after a 2024 extension, and its 2026 data-centric AI stack targets Level 4 autonomous driving under limited conditions.
Turing
Turing, founded in Tokyo in 2021, is developing end-to-end autonomous-driving AI using vision-language-action models, which combine visual perception, language understanding and action to turn camera footage into driving decisions.
Backed by Denso and JIC Venture Growth Investments, the startup pulled in ¥15.27 billion for its Series A first close in November 2025, splitting the capital between ¥9.77 billion in preferred equity and ¥5.5 billion in syndicated loans.
While TIER IV focuses on building a shared autonomy platform, Turing favours a vertically integrated approach built around a single AI model.
SkyDrive
SkyDrive, founded in Tokyo in 2018, is one of Japan’s leading electric vertical-takeoff-and-landing aircraft developers. It works on the SD-05 eVTOL aircraft and heavy-lift drones.
Its relevance to physical AI is in the certification, safety systems, sensing and flight-control software that automated aerial mobility depends on. It received a regulatory certification basis in Japan in 2025 and counts MUFG, Suzuki, JR East and NEC among its investors, a list that spans finance, carmaking, rail and electronics.
ugo
Tokyo-based ugo, founded in 2018, builds semi-humanoid service robots for security, inspection, guidance and facility monitoring.
The company introduced ugo Nova, a semi-humanoid system with dual compliant arms and an omnidirectional chassis, to support Japan’s AIRoA initiative. The project operates under a ¥20.5 billion NEDO contract with plans to roll out more than 100 domestic dual-arm mobile machines across retail shelf restocking, warehouse parcel sorting and automotive parts kitting.
It’s one of the Japanese companies selected to develop standardised dual-arm work robots that will collect data for physical AI.
Sensyn Robotics
Tokyo-based Sensyn Robotics develops AI-powered robots for inspection and maintenance in industrial environments.
Keeping facilities and critical systems running safely is the less flashy, commercially vital backbone of physical AI. Infrastructure maintenance is a clear priority in the national strategy alongside factory automation, mobility, logistics and healthcare, which puts Sensyn squarely in the physical AI conversation.
