CES 2026 LG CNS CEO says industrial robots to enter production lines within 2 years
LG CNS CEO Hyun Shin-gyoon speaks during an interview on the sidelines of CES 2026 in Las Vegas, Jan. 7. Courtesy of LG CNS
Mass production, not technology, to drive commercialization timeline
LAS VEGAS — LG CNS CEO Hyun Shin-gyoon said industrial robots are expected to soon move beyond proof-of-concept (PoC) stages and begin working on production line.
“Moving from PoC to a stage where robots actually work on production lines will take around two years,” Hyun said during an interview with reporters on the sidelines of CES 2026 on Jan. 7 (local time).
He added that LG CNS is not pursuing competition in robot hardware but is instead focused on training and operating robots so they can function in real industrial environments.
Hyun said that although robot technologies themselves are already close to commercialization, deploying robots in real workplaces requires additional preparation.
“Showing robots moving and having them work alongside people on production lines to improve productivity are completely different matters,” he said, stressing that processes, integration with workers and operational environments must be prepared together.
Hyun said the two-year timeline is based on the time required to establish mass production systems rather than technological uncertainty.
“For robots to become price-competitive, a mass production system is necessary,” he said.
According to Hyun, building robot production lines typically takes more than a year, followed by another year to stabilize component supply chains. Once those conditions are met, he said, robots can be deployed more widely in industrial sites.
While robots such as wheeled platforms and fixed industrial robots are already being used at scale, Hyun said bipedal and quadruped robots are expected to follow as production systems mature.
Hyun defined LG CNS’s role in the robot ecosystem as an on-site application specialist rather than a hardware manufacturer. He divided the robot industry into three areas — hardware manufacturing, development of general-purpose “brains” and on-site application — and said LG CNS focuses on the third.
“Even if robots are equipped with general-purpose brains, they cannot be used immediately in the field,” Hyun said. “Someone has to collect site data, retrain the robots, operate and monitor them and retrain them again if performance declines. That is the role LG CNS plays.”
LG CNS is working with Chinese robot hardware maker Unitree and is using robot foundation models developed by U.S.-based Skild AI. Hyun said the company is pursuing a hardware-neutral strategy and does not depend on any single robot manufacturer or artificial intelligence (AI) model.
Hyun said that in the era of physical AI, general-purpose intelligence in software will become standardized more quickly. He compared the trend to the large language model market, where performance gaps narrowed rapidly.
“Ultimately, competitiveness will depend on how well AI is applied and operated in real industrial environments,” he said.
LG CNS is linking its robot strategy to its AI transformation initiative, aimed at improving efficiency within group affiliates and enhancing products and services through AI.
“No matter how many robots are produced, expansion will be limited without companies that can select, train and manage them in the field,” Hyun said.
“LG CNS aims to become a company that makes robots actually work.”