KAIST signs MOU with 16 companies at “Physical AI Tech Week”
AI chips, digital twins, and robotics technologies under one roof

The Korea Advanced Institute of Science and Technology (KAIST) is establishing a cooperation framework with 16 domestic physical artificial intelligence (AI) companies for technology standardization and joint expansion into overseas markets. The agreement is part of the activities of the Governance Subcommittee of the Ministry of Science and ICT’s Physical AI Alliance, chaired by KAIST Professor Jang Young-jae, and aims to open up global markets by linking the robotics, AI, and manufacturing system technologies held by individual companies.

On the 13th, KAIST signed a memorandum of understanding (MOU) with the 16 companies to “build a physical AI standardization cooperation framework” at “Physical AI Tech Week,” held at COEX in Seoul.

Representatives of KAIST and 16 domestic physical AI companies pose for a commemorative photo after signing an MOU to 

“build a physical AI standardization cooperation framework” at “Physical AI Tech Week,” held at COEX in Seoul on the 13th. / Reporter Lee Yun-jeong

Participating physical AI companies included Megazone Cloud, Doosan Robotics, Rebellions, Shinsung E&G, T-Robotics, Green Power, MX Robotics, Pebblous, Persona AI, Cantops, Omelet, Atec Mobility, Movensys, and Navifra.

The agreement was designed to expand the use of KAIROS, a robot orchestration platform developed by KAIST, and to pursue technology standardization and joint overseas expansion, centered on the Physical AI Alliance’s Governance Subcommittee. KAIROS is a physical AI-based unmanned factory testbed that integrates control of various robots, equipment, and AI to operate an entire factory as a single AI system.

“Private companies came together to pursue this MOU with the aim of not stopping at building the platform, but exporting together based on it,” explained Jang Young-jae, KAIST professor and chair of the Physical AI Alliance’s Governance Subcommittee.

He went on to stress the importance of companies not simply developing AI, robots, and manufacturing systems separately, but exporting them as integrated solutions. This agreement is the first round of MOUs, and plans to expand the number of participating companies will also be discussed going forward.

Three subcommittees covering everything from technology development to industrial demonstration and standardization

The agreement was carried out as part of the activities of the second Physical AI Alliance, launched in June.

The Ministry of Science and ICT’s Physical AI Alliance is a private sector-led consortium that brings together the capabilities of industry, academia, and research institutes to advance physical AI technology development, industrial deployment, and ecosystem building. Cho Jun-hee, chairman of the Korea Artificial Intelligence & Software Industry Association, serves as co-chair.

The alliance consists of three subcommittees: Technology Full Stack, Industry Bridge, and Governance.

The Technology Full Stack Subcommittee works to secure K-Physical AI full-stack technology that combines domestic AI models, AI chips, software, and robotics. The Industry Bridge Subcommittee connects demand-side and supply-side companies in industries such as manufacturing, defense, and maritime to identify demonstration and commercialization projects.

The Governance Subcommittee, which drove this MOU, is responsible for standardizing the physical AI ecosystem and building a global cooperation framework. Its key tasks include interoperability between diverse manufacturing equipment and heterogeneous robots, data standardization, establishing a physical AI education system, and joint overseas expansion.

Jang emphasized that for the physical AI industry to grow, a wide range of equipment and components must be able to operate according to common standards. He also explained that private companies need to respond jointly when it comes to developing global markets and training talent.

In particular, he said that while standardization and design automation of physical AI-based factory design technology are necessary, the human role cannot be eliminated in that process, so related education must be pursued in parallel.

Jang said he plans to expand cooperation between companies and continue follow-up discussions on standardization and global market entry. Specific plans will be announced through official channels.

From AI chips to robots… physical AI technologies on display

The venue also featured an exhibition space showcasing the core technologies that make up physical AI, including AI chips, data and digital twins, manufacturing AI, and robotics, along with cases of their application in industrial settings.

Companies and research institutes including Rebellions, FuriosaAI, Megazone, Neuromeka, Nota AI, RLWRLD, KT, and LG CNS participated, presenting physical AI technologies and real-world industrial applications.

The exhibition introduced a variety of technologies through physical products, videos, and live demonstrations, including neural processing units (NPUs) for AI inference in data centers and edge environments, manufacturing AI-based virtual factory control and optimization simulations, intelligent robots and humanoids, and micro edge servers.

Bae Kyung-hoon, Deputy Prime Minister and Minister of Science and ICT, also toured the exhibition to review the physical AI technologies and business status of major companies.

Meanwhile, “Physical AI Tech Week,” hosted by the Ministry of Science and ICT and organized by the Physical AI Alliance and the National IT Industry Promotion Agency (NIPA), runs for two days, the 13th and 14th, at COEX in Seoul. On the first day, “Physical AI Alliance Day,” the government’s 2027 physical AI program direction and each subcommittee’s cooperation projects were shared. On the 14th, “Physical AI Tech Day” will be held under the theme “Intelligence into Action.” Keynote speeches by domestic and international experts, technical presentations, a talk concert, and a technology exhibition will share the latest technologies and industry trends.

Source: https://it.chosun.com/news/articleView.html?idxno=2023092171947