Polaris AI Strengthens Humanoid Technology Collaboration with UBTECH... "Advancing Korean-Style Physical AI Models"

The training session on WalkerS2 operation and SDK-based secondary development held at Polaris AI headquarters. Provided by Polaris AI

Polaris AI has begun strengthening its domestic physical AI commercialization capabilities through technological collaboration with global robotics company UBTECH Robotics. The company announced that it is conducting technology training related to the industrial humanoid robot ‘Walker S2,’ and is specifically evaluating the feasibility of applying it to actual manufacturing sites.

On May 28, Polaris AI reported that it recently conducted operational training for the Walker S2 and SDK (Software Development Kit)-based secondary development sessions together with engineers from UBTECH headquarters. During this training, the company received the development environment and coding files required for robot motion control, and was also trained in core motion control technologies, including hand joint movement.

The company explained that this training is significant because it goes beyond a simple product introduction and has established a foundation for implementing and adjusting work scenarios required in actual industrial environments, all based on software.

It is known that the company directly verified basic tasks such as box handling, and reviewed scenarios that could be applied to a variety of industrial settings, including logistics, manufacturing, inspection, and assembly.

The Walker S2 is a next-generation humanoid robot developed by UBTECH for use in industrial sites. According to the company’s official materials, it features autonomous battery replacement and can handle work loads of up to 15 kilograms. Its potential applications are attracting attention in smart factory and manufacturing process-centric industrial environments. Recently, discussions and demonstration cases for the adoption of humanoid robots have been rapidly expanding in the global market, particularly in advanced manufacturing industries such as automotive, aerospace, and logistics.

Polaris AI plans to use this technology training as a starting point to gradually develop humanoid robot application models tailored to the domestic manufacturing environment. The company is emphasizing its strength in having manufacturing-based affiliates such as Polaris Sewon, Polaris Uno, and Polaris AI Pharma.

The company intends to identify a variety of work scenarios based on its affiliates, which have different process structures in fields such as automotive parts and materials, as well as bio and healthcare manufacturing. These scenarios include parts transport, repetitive assembly, packaging, inspection assistance, quality control, and equipment inspection. The company aims to expand these into customized physical AI solutions for each client in the future.

Furthermore, the company is also focusing on the fact that implementing humanoid robots in industrial sites requires more than just hardware provision; it also needs work environment analysis, motion scenario design, SDK-based control, client-specific customization, and maintenance systems. Polaris Group plans to prioritize acquiring software control and field application capabilities as it builds a Korean-style physical AI business model.

A company representative stated, “This training is the first practical step toward understanding and applying humanoid robots to the domestic industrial environment. By combining UBTECH’s robotics hardware technology with Polaris Group’s AI and software capabilities, we will develop physical AI models that can be utilized at real manufacturing sites.”

The representative added, “Since the humanoid robot market is still in its early stages, it is more important to steadily build up field application capabilities rather than making exaggerated adoption announcements. We plan to expand our business in a direction that improves productivity in the domestic manufacturing and logistics industries through SDK-based secondary development, work scenario design, and client-specific application reviews.”

Meanwhile, Polaris AI is also working to strengthen its business foundation by recently updating its robotics business website and expanding its participation in the physical AI industry network. Going forward, the company plans to link humanoid robots, vision AI, and industrial automation software to identify substantial business opportunities within the domestic physical AI ecosystem.

This technology training is seen as a move showing that Polaris AI is actively moving beyond simple robot distribution and is now working in earnest to build Korean-style physical AI solutions applicable to real industrial sites.

This content was produced with the assistance of AI translation services.