A subsidiary of LG has showcased a future logistics workflow in which multiple autonomous robots collaborate without human intervention. In a demonstration, a bipedal humanoid lifted a box from a conveyor belt and passed it to a wheeled quadruped transport robot, which autonomously delivered it across the site.
A wheel-type humanoid then used its extended arms to place the box on a shelf over two meters high. The entire process was powered by LG CNS’s Physical Works platform, with robots independently recognizing objects, making decisions, and coordinating tasks using Robot Foundation Model technology.
Recently, LG Display also unveiled a 7.2-inch curved P-OLED humanoid robot display powered by its third-generation Tandem OLED display technology.
Robots run logistics
LG CNS showcased its PhysicalWorks robot transformation (RX) platform through a live logistics demonstration at its Magok campus in western Seoul. During the event, four robots from different manufacturers completed coordinated warehouse tasks without remote control. The platform combines a simulation and video-based robot training module with a real-time orchestration system capable of assigning and reassigning jobs across mixed-brand robot fleets, according to The Korea Herald (TKH).
During the demonstration, a humanoid robot picked up packaged goods from a conveyor and loaded them into a box. A wheeled quadruped logistics robot then transported the box across the workspace to a wheeled humanoid robot, which stacked it on a designated shelf. After completing the delivery, the quadruped returned to continue the logistics cycle while the wheeled humanoid placed an empty box back onto the conveyor for refilling.
LG CNS also demonstrated adaptive task management by simulating an emergency scenario. When the quadruped robot was reassigned to patrol operations, the platform automatically deployed another autonomous logistics robot to continue the transport work without interrupting the workflow, according to TKH.
The company said the robots operated entirely autonomously, relying on the platform’s robot learning and coordination capabilities rather than manual intervention. In the live setup, the robots transferred a single box between stations positioned roughly two to three meters ((6.5 to 9.8 feet) apart in about 90 seconds, with performance expected to improve further through additional field training and operational experience.
Intelligent fleet coordination
LG CNS said the robotics industry is shifting beyond hardware-focused development toward software systems that allow robots to understand, coordinate, and reliably execute real-world industrial tasks. The company believes successful robot transformation depends on integrated learning, verification, and operational frameworks tailored to manufacturing and logistics sites rather than on individual robot performance alone, reports Chosun Biz.
To support this strategy, LG CNS has developed its Physical Works platform, which combines two systems: Physical Works Forge for robot data learning and simulation-based verification, and Physical Works Baton for centralized control of robots from multiple manufacturers. The platform supports different robot types, including bipedal humanoids, quadrupeds, wheel-based robots, AMRs, and AGVs, through a unified management interface.
LG CNS said the system reduces robot deployment timelines from several months to around one or two months. In mixed-robot environments of about 100 units, the company projects productivity gains exceeding 15 percent and operating cost reductions of up to 18 percent by minimizing traffic overlap, congestion, and manual intervention, Chosun Biz reported.
The company has also invested in robotics and embodied AI firms developing humanoid control and robot foundation models. Currently, LG CNS is conducting proof-of-concept projects with 20 customers across industrial sectors while deploying the platform in South Korea’s Busan Smart City pilot project to manage patrol, cleaning, delivery, and service robots through a single system.