[No-Filter Robot] Soccer-Playing Robots, Thinking Factories... The Future of K-Manufacturing Armed with 'AX'

Visiting the 2026 Smart Factory & Automation Industry Exhibition (AW 2026)

From Humanoids to Robot Hands: Closing the 'Dexterity Gap' and Ushering in the Era of Autonomous Manufacturing

Strength of Robust Software Revealed, but the Slowing S

On March 4, at the "2026 Smart Factory and Automation Industry Exhibition (AW 2026)" held at COEX in Gangnam-gu, Seoul, visitors were greeted not by the heavy sounds of machinery but by the cheerful laughter echoing through the venue. The source of this sound was the booth of TopRobot, a humanoid rental company located at the entrance to Hall B on the first floor. There, a 'little robot' about the size of a five-year-old child was chasing a red toy soccer ball. When a company staff member playfully tried to take the ball away, the robot dashed forward like a real child, and, driven by enthusiasm, repeatedly missed kicks, narrowly avoiding falling on its backside.

A visitor at the event commented, "I was expecting the robot to dribble like a soccer prodigy, but I was surprised it couldn't kick the ball properly," adding, "When the ball disappeared from its view, the robot looked around and cocked its head, missing its kicks, which made it seem just like watching a real child. It was fascinating." This was not just a machine following a pre-programmed trajectory but a scene that symbolized the evolution of physical AI, capable of responding to its surroundings in real time.

On the 4th, at COEX in Gangnam-gu, Seoul, a small humanoid robot is playing ball at the "2026 Smart Factory and Automation Industry Exhibition (AW 2026)." Photo by Jin Young Kim

All Eyes on 'Atlas'... A Fierce Showdown Among the Industry's Unsung Robotics Champions

This adorable soccer prodigy robot was merely a preview of the exhibition's main theme: manufacturing 'AX' (AI Transformation). As visitors entered Hall D, the main stage of the exhibition, the present and future of Korean robotics came into full view. The joint booth of Hyundai Glovis and Hyundai Robotics Lab, occupying an entire wall, was undoubtedly the landmark of this exhibition. In particular, a non-powered model of Boston Dynamics' "Atlas," unveiled for the first time in Korea, drew such a crowd eager to see the star of "CES 2026" in person that there was barely space to move.

Lee Kyubok, President of Hyundai Glovis, who attended the event, stated, "Atlas is currently being trained first at Hyundai Glovis' integrated logistics center at the Savannah plant (HMGMA) in Georgia, USA," adding, "We plan to deploy it first to advanced logistics operations in 2028 and then to manufacturing processes by 2030."

Hyundai Glovis' Ziku Shuttle is performing logistics operations. Photo by Kim Jinyoung

For heavy cargo that Atlas could not lift, the task fell to the Ziku Shuttle. Known as a "palette shuttle," this equipment automatically transported pallets loaded with transport robots along fixed rails, transferring goods weighing more than two tons to inbound, storage, and outbound locations. The warehouse control system (WCS) platform "Orca," co-developed by Hyundai Glovis and its subsidiary Altiol, analyzes the different sizes and internal structures of each warehouse to set optimized routes.

The stars of AW 2026 were not limited to large corporations. So-called "unsung champions" of robotics, such as Yujin Robot and Goseong Engineering, showcased their strengths by deploying specialized autonomous mobile robots (AMRs) and collaborative robots, demonstrating the robust backbone of the Korean robotics ecosystem. RS Automation, specializing in robot motion control technology, reinvented a delta robot optimized for pick-and-place tasks as a claw machine, attracting many visitors.

A demonstration of a product from Goseong Engineering, a partner of Mobed developed by Hyundai Motor Group Robotics Lab. Photo by Jinyoung Kim

Robotic Hands Narrow the 'Dexterity Gap'... "Challenging the Realm of Humans"

One particularly impressive aspect of this exhibition was the focus of many companies on "robot hand" technology. Numerous robotics companies, including Robotis, Wonik Robotics, and Global Zeus, prominently displayed robotic arms that accurately mimicked human hand functions. Some robots delicately transferred soft fruits such as peaches and persimmons to plates without squashing them, while others, like Wonik Robotics' "Allegro Hand," spun strawberries and plums with their fingertips, as if about to pop them into someone's mouth. Whereas traditional robots focused on "strength" to move heavy objects, the robots of 2026 emphasized "delicacy," skillfully picking up small and soft fruit models without crushing them.

Wonik Robotics' Allegro Hand grasping a fruit model and rotating it with fingertips. Photo by Jinyoung Kim

Ken Goldberg, a renowned roboticist and professor in the Department of Industrial Engineering at UC Berkeley, believes that there are still engineering challenges for robots to replicate humans' unconscious fine finger movements or adjust grip strength based on the texture of objects. In a recent interview with this newspaper, he noted, "It will take several more years before robots acquire the complex manipulation abilities of electricians, mechanics, or plumbers," adding that it is still at least a decade too soon for robots to replace human labor.

However, the robotic hands showcased at the event vividly demonstrated how the technological barrier of the "dexterity gap" — the limitation of robots to match the precision of human hand movements pointed out by Professor Goldberg — is being overcome. Robotic hands integrated with AI can now identify the rigidity and shape of objects, autonomously determining the optimal gripping method. This moment offered a glimpse into the future of "autonomous manufacturing," where AI goes beyond simple automation to become the brain and muscle of the factory, allowing facilities to make and execute decisions independently.

A robot from Global Zeus is transferring fruits onto a plate. Photo by Jin Young Kim

The Command Center of AI Manufacturing Sites... The Thinking Factory 

While robots dazzled with their performances, companies positioning themselves as the "brains" of the manufacturing field, such as LS Electric and Schneider Electric, showed a strong presence elsewhere on the exhibition floor. LS Electric, in particular, which had the largest booth at this exhibition, placed its new PLC product, the "SU-CM70," at the forefront, presenting the future of industrial automation solutions integrated with its AI factory model. By connecting virtual factories to real sites via software (digital twin), AI provided real-time feedback on which processes had reduced production efficiency and where defects were occurring. This is the birth of the "thinking factory," enabling robots seen earlier to demonstrate their full capabilities in industrial settings.

LS Electric's AI Factory collecting equipment data from semiconductor plants and detecting hazards in industrial sites. Photo by Jinyoung Kim

This AW2026 signaled that "K-manufacturing" is entering an era of "autonomous manufacturing," where factories independently assess situations and produce optimal products with minimal human intervention. AI technology, which provides a comprehensive view of factory data flows and maximizes energy efficiency, is becoming essential infrastructure for survival, not just a passing trend.

However, beneath the spectacle of advanced technologies, challenges remain. There are concerns that the technological gap between large enterprises and small and medium-sized businesses in the manufacturing sector will continue to widen. In fact, since 2019, the government has supported AX for small and medium-sized enterprises by establishing the Smart Manufacturing Innovation Promotion Group under the Ministry of SMEs and Startups, but the annual adoption of smart factories has recently slowed. According to the "First Survey on the Status of Smart Manufacturing Innovation" released last year, the adoption rate of intelligent (smart) factories among small and medium-sized enterprises is about 18.6 percent — only around one-fifth that of medium-sized companies (85.7 percent).

A company representative met at the event explained, "Large companies have the capacity to realize AI transformation on their own, but small and medium-sized enterprises, who are the main customers for industrial automation solutions, often hesitate to invest due to the initial cost burden," stressing the need for government support to ensure the growth and mutual prosperity of the autonomous manufacturing industry.

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