A humanoid packed the box. A “robot dog” carried it away. Then a wheeled robot lifted it onto a shelf higher than most people can reach.
LG CNS, the IT services arm of South Korea’s LG Group, demonstrated a multi-robot warehouse system at LG Science Park in western Seoul, showing humanoids, quadruped robots, autonomous mobile robots, and automated guided vehicles working together without human remote control.
The platform, called PhysicalWorks Baton, allows humanoid robots, quadruped robots, autonomous mobile robots (AMRs), and automated guided vehicles (AGVs) to work together through a single management system. LG CNS said the goal is to simplify the way companies deploy and manage large robot fleets in warehouses and factories.
During the demonstration, a bipedal humanoid robot from China’s Unitree packed items into a box before passing the load to Deep Robotics’ quadruped robot, M20. The robotic carrier transported the box across the site to Dexmate’s wheeled humanoid robot, Vega, which placed it onto a shelf more than two meters high.
After the delivery, the robots reset the workflow themselves, returning empty containers to the conveyor line to continue the cycle.
Robots reacted to an emergency on their own
LG CNS also staged an emergency scenario to show how the system could adapt in real time. According to Korea JoongAng Daily, a warning alert suddenly appeared during the demonstration, prompting the quadruped robot to abandon its delivery task and switch to patrol duties. Another logistics robot, Bear Robotics’ Carti-100, automatically took over the transport work, allowing operations to continue without interruption.
“The operating platform monitors where each robot is and what it’s doing, then centrally manages the task order to distribute work,” an LG CNS official said, according to Korea JoongAng Daily. “Robots from different manufacturers and of different kinds can adapt, switch roles and collaborate as needed.”
LG CNS said none of the robots in the demonstration were remotely controlled by humans. The machines recognized objects, coordinated movements, and reassigned jobs autonomously through what the company described as Robot Foundation Model technology.
Faster deployment, lower costs
LG CNS said the platform could significantly reduce the time needed to deploy industrial robots.
The company estimates that PhysicalWorks Baton can cut deployment timelines from several months down to one or two months. In environments operating with around 100 robots, LG CNS projects productivity gains of more than 15% and reductions in operating costs of up to 18% by minimizing congestion and improving coordination.
While the warehouse demo was a proof of concept, the technology is already hitting the real world. LG CNS is currently running pilot projects with 20 customers across the electronics, chemical, and shipbuilding industries. In shipyards, the robots are even being tested for welding tasks to help solve labor shortages.
Beyond the factory floor, the Baton system is already active in South Korea’s Busan Smart City project. There, it manages a diverse fleet of barista, cleaning, patrol, and porter robots.
“We are first applying the data, which we have accumulated over the years through our IT services, to logistics and manufacturing environments,” said Lee Jun-ho, senior vice president of LG CNS’ smart logistics & city business division, according to the Korea JoongAng Daily. “We expect to see solid results within two years.”
Also read: South Korea’s robot monk Gabi joined a Buddhist ceremony, raising new questions about AI, religion, and cultural tradition.