scorecardresearch

Friday, November 6, 2026

Support our Journalism

Depth, context and analysis

Subscribe
HomeThePrint EssentialSouth Korea has built a factory to teach humanoid robots how to...

South Korea has built a factory to teach humanoid robots how to do household chores

South Korea’s first AI Humanoid Data Factory will use 12 robots to collect real-world task data.

Stay updated with ThePrint:

New Delhi: A robot taking clothes out of a washing machine, picking groceries from a refrigerator or operating the buttons inside a car may look like a glimpse of a future where machines handle everyday chores. South Korea is building the infrastructure to make that future possible, with a new facility designed to teach humanoid robots how to perform tasks in the real world.

The country inaugurated its first dedicated AI Humanoid Data Factory on 3 November at the Korea Institute of Machinery and Materials (KIMM). The facility will collect data from robots performing everyday and industrial tasks, using it to train artificial intelligence (AI) models to understand human instructions, navigate their surroundings and eventually operate autonomously.

The factory currently houses 12 humanoid robots, which are remotely controlled by human operators as they carry out different activities. Their movements and task-related information are recorded to create datasets for training AI systems. KIMM plans to gradually increase the number of robots and expand the range of tasks they perform.

At the inauguration, robots demonstrated household activities such as taking items out of a refrigerator, handling laundry and arranging a sofa. They also performed industrial tasks, including operating buttons and levers inside a vehicle.

Teaching robots through real-world tasks

The facility has been established by KIMM in collaboration with the Electronics and Telecommunications Research Institute (ETRI), under South Korea’s Global TOP Strategic Research Group for Self-Growing AI Humanoids.

The data collected will feed into a repeated training process. AI models trained on these datasets will be deployed on physical robots and their performance evaluated. Data from these trials will then be fed back into subsequent rounds of training, allowing the models to improve through repeated interactions with real-world environments.

The aim is to develop AI systems that can be deployed across different humanoid robot platforms, including South Korea’s K-Humanoid standard platform. The inauguration also featured demonstrations by KAIROS, a Korean-developed humanoid robot, which performed precise object manipulation and full-body motion imitation.

South Korea’s push for homegrown humanoid AI

The initiative is part of South Korea’s effort to build domestic capabilities in humanoid robotics and reduce dependence on overseas data and AI models. By creating infrastructure to collect large volumes of task-specific data within the country, the government hopes to strengthen its technological self-reliance and domestic robotics industry.

KIMM president Seog-Hyeon Ryu said access to large volumes of high-quality robot task data was essential to national technological competitiveness.

The institute plans to gradually open the facility to domestic companies, universities and research institutions from the end of 2027, allowing them to use the infrastructure for research and development.

The Ministry of Science and ICT, which supports the initiative, said it would continue bringing together the country’s research capabilities under the K-Moonshot AI Humanoid Mission. The programme aims to develop independent humanoid AI technologies and strengthen the competitiveness of South Korea’s domestic industry.

(Edited By Raunaq Lekhi)

Subscribe to our channels on YouTube, Telegram & WhatsApp

For readers who want the full picture.

You read our articles closely. Subscribe to go further, with exclusive newsletters in your inbox, flagship events hosted by Shekhar Gupta, AMAs with our senior editors, the chance to write under your own byline, and more.

Subscribe to ThePrint →

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Popular