As robotics evolved, the industry has steadily adopted more robots to automate many kinds of runs. In 2023, more than 540,000 new industrial robots were installed worldwide, acquiring more than 4 million industrial robots per IFR.
Industrial robots are generally difficult to perform excellent but accurate work, handle delicate materials, and adjust them according to changing conditions. Robots in a restaurant kitchen are helpful in more ways than for example. That’s why many industrial processes are still manuals.
Korea Startup RLWRLD aims to solve this problem of basic AI models specially built for robotics by combining large language models with traditional robot software. The company says this model allows robots to create fast and agile movements and perform some “logical reasoning.”
Jung-HEE Ryu, the founder and CEO of RLWRLD, said, “You can use the RLWRLD’s basic model to learn a process that requires a lot of manual tasks to learn human expertise and copy to create a more efficient work environment.
Startups are now coming from stealth with seed funds of 21 billion KRW (about $ 14.8 million). The round was led by the venture capital company, Hashed, and also invested Mirae Asset Venture Investment and Global Brain.
In particular, RLWRLD has attracted large strategic investors such as Japan’s ANA Group, PKSha, Mitsui Chemical, Shimadzu and KDDI. LG Electronics and SK Communications in Korea; Amber in India.
RLWRLD said that seed funds will be used to fund the concept proof projects with strategic investors. Security computing infrastructure, such as a device that collects GPUs, purchasing robots and a wide range of data; Hire the best research talent. Startup also said he would use new money to develop advanced hand movements associated with five fingers, Ryu said that competitors such as Tesla, Picture AI and 1X are not yet proven.
RYU said RLWRLD is working with strategic investors to use the AI model to seek ways to automate various human -centered workflows. They are preparing for humanoid -based autonomous action demonstrations scheduled later this year, Ryu said. The company is also trying to develop a platform that can support various types of robots, including industry, collaboration, autonomous mobile robots and humanoids.
Founded in 2024, RLWRLD is Ryu’s third startup. His second start -up, Olaworks, was acquired by Intel in 2012 and eventually became the Korea R & D center of Intel in the computer vision department. And in 2015, he founded Future Play, a startup accelerator focusing on deep sea technology companies.
Ryu said that when it inspired to start a new company, I knew how quickly AI startups were increasing in the US, Europe and China, and that similar AI new companies in Korea and Japan were relatively absent.
He talked with more than 30 AI professors in Korea and Japan about the lack of infrastructure such as data and GPUs and the discouraged obstacles and available opportunities to start ventures.
“I decided to prioritize the Robot Engineering Foundation model (RFM) for the technical saturated LLM field.
Since then, he has brought six professors from Korea’s top institutions to launch RLWRLD with the research team, including KAIST, SNU and POSTECH.
RLWRLD is not alone in solving this problem. New companies, such as Skild AI and Physical Intelligence, are building a basic model similar to robotics, just like large companies such as Tesla, Google Deepmind and NVIDIA.
However, Ryu thinks that his startup already has AI and robotics experts, so he has a good start to develop humanoid robots with high levels of freedom as well as basic models for robotics.
“In addition, such a company generally relies on low coating robots such as two finger gripper. RLWRLD has already secured a high application reference robot, so we expect excellent performance results.”
Ryu also said that thanks to strategic investors, RLWRLD can quickly collect valuable data from a nearby manufacturing site. In 2024, Japan and Korea accounted for 9.2%of the world’s manufacturing production.
RLWRLD aims to generate profits through a joint concept proof (POC) project and collaboration demonstration with strategic partners earlier this year.
The long -term goal of the startup is to accept robots that can be used in a domestic environment to help factories, distribution centers and retail stores and household chores. In the meantime, the priority is for the industry because the industry pays the most cost and has a strong demand for automation.
Startups have 13 employees.
