AIToday
RoboticsAI Business & IndustryImpress WatchPublished: Oct 27, 2026, 13:00 JST

KDDI, ELYZA launch ELYZA Robotics, 30 billion yen into physical AI

Send on LINE
KDDI, ELYZA launch ELYZA Robotics, 30 billion yen into physical AI

3 Key Points

  1. What happened

    KDDI and ELYZA set up ELYZA Robotics and began operations on October 1; the KDDI group will put 30 billion yen into R&D through fiscal 2030 and aims to deploy 1,000 or more robots in the field.

  2. Why it matters

    The venture bundles ELYZA and KDDI Research talent into one organization of more than 300 specialists, targeting tasks in retail, logistics and manufacturing through KDDI group companies such as Lawson and Soracom.

  3. What to watch

    Proof-of-concept work is scheduled to begin in 2026 across a wide range of sites, using industrial, quadruped and humanoid robots procured from Japanese and overseas vendors.

WHO IT HITSField operators in retail, logistics and manufacturing are the first likely candidates to see these robots, since shelf stocking and inventory, transport and shortage detection, and precision work are the named use cases. Robotics vendors supplying industrial, quadruped and humanoid machines stand to gain procurement orders as proof-of-concept work scales.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

ELYZA Robotics combines the AI research side of ELYZA with KDDI's communications network, its "Digital Belt" AI computing infrastructure and remote monitoring sites. The company is set up to handle everything from research and development to social implementation, organized around three pillars: robot foundation models, embedding robots into field operations, and a physical AI platform.

On the technology side, the work covers VLM models that jointly understand images and language and VLA models that output robot actions, plus techniques for capturing work data recorded from a human viewpoint and reproducing sites in a simulator. ELYZA Robotics plans two versions of its physical AI platform, one for R&D that supports foundation model development and the accumulation and use of field data, and one for actual use that supports distributed inference, remote monitoring and coordinated work by multiple robots, feeding field data and know-how back into research.

Deployment is set to start with proof-of-concept work in 2026, procuring robots from domestic and overseas vendors and working with KDDI group companies including Lawson and Soracom as well as partner firms. The use cases are concrete: shelf inventory and stocking and customer service in retail, transport and shortage detection in logistics, and precision work in manufacturing. Agriculture and construction are named as future expansion areas, which suggests the company is positioning physical AI as a long-term business rather than a single-product bet.

FAQ
When does ELYZA Robotics start operating?
The new company began business on October 1, with its head office in Bunkyo, Tokyo, and capital of 100 million yen.
Which robots will the company use?
It plans to procure industrial robots, quadruped robots and humanoid robots from domestic and overseas robot vendors and offer them by industry and use case.
Where will the robots be tried first?
Proof-of-concept work starts in 2026, with retail, logistics and manufacturing named as initial areas and agriculture and construction as future targets.
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleAnthropic expands AI support for nonprofits as it files to sell shares to the public