South Korean IT services company LG CNS is stepping deeper into the industrial robotics market with a new platform designed to dramatically shorten the time required to deploy robots in real-world manufacturing environments.
At a media event held at LG Science Park in Seoul on Wednesday, LG CNS introduced “PhysicalWorks,” an integrated robotics transformation (RX) platform that manages the entire lifecycle of industrial robots — from data collection and AI training to deployment, operation, and centralized control.
The company also demonstrated multiple types of robots autonomously collaborating in a logistics scenario without direct human remote control, marking what LG CNS described as a first in South Korea for heterogeneous robot cooperation.
Industry observers say the announcement reflects a broader shift in manufacturing automation, where companies are increasingly looking beyond simple robot installation and focusing instead on AI-driven robotic ecosystems capable of learning, adapting, and operating continuously in changing industrial environments.
One of the biggest barriers to industrial robot adoption has traditionally been deployment speed. Configuring and optimizing robots for factory operations often takes several months due to complex training, testing, and workflow integration processes.
LG CNS says its new platform could reduce that process to one or two months.
The company’s “PhysicalWorks Forge” platform focuses on robot training and simulation. The system integrates data collection, AI learning, testing, and field optimization into a single workflow intended to streamline robot deployment.
Rather than relying solely on repetitive physical demonstrations by human operators, the platform uses 3D virtual simulations of industrial environments to generate training data more efficiently. LG CNS said future upgrades may also include motion-capture systems and AI learning based on human work videos.
The platform also automates portions of the data refinement process. AI systems selectively identify useful training behaviors while filtering out unnecessary or failed movements to improve learning efficiency.
Once training is completed, robots are tested in virtual simulation environments before being optimized for actual industrial deployment. Analysts say this type of simulation-based robotics training is becoming increasingly important as factories seek to scale automation without lengthy downtime or manual recalibration.
LG CNS also introduced “PhysicalWorks Baton,” a unified robot orchestration and control platform designed to manage multiple robot types simultaneously. The system supports humanoid robots, quadruped robots, wheeled robots, autonomous mobile robots (AMRs), and automated guided vehicles (AGVs) within a single operational framework.
One of the major challenges in industrial robotics has been fragmentation among robot manufacturers, with different control systems and operational interfaces creating management complexity. LG CNS said Baton standardizes robot operational data and control systems, allowing robots from different manufacturers to work together more efficiently.
The platform also incorporates agentic AI capabilities that dynamically respond to unexpected operational changes. If a conveyor belt stops or a robot becomes unavailable, the system can automatically redistribute tasks and redesign logistics routes in real time.
Industry analysts say such adaptive coordination systems are becoming critical as factories move toward fully autonomous operations and “lights-out” manufacturing environments.
According to LG CNS, applying PhysicalWorks Baton to a large-scale environment with around 100 robots could improve productivity by more than 15% while reducing operational costs by up to 18%.
The company said the strongest gains are expected in facilities operating mixed fleets of robots from different manufacturers, where traffic congestion, redundant movement, and manual intervention have historically reduced efficiency.
A key feature of LG CNS’s strategy is the integration between the Forge and Baton systems. Data collected during real-world robot operations can be continuously fed back into training environments, creating a self-improving cycle of learning, validation, and operational optimization.
During the event, LG CNS demonstrated four robot types — including humanoid, quadruped, wheeled, and autonomous logistics robots — working together autonomously in a logistics workflow without remote human control.
The company said its Forge platform is currently involved in proof-of-concept projects with more than 20 clients, while Baton has already been deployed in South Korea’s Busan Smart City pilot project, where it manages patrol, cleaning, delivery, and barista robots within a unified control system.
Industry experts say the growing convergence of AI, robotics, simulation, and autonomous coordination technologies could significantly accelerate adoption of physical AI systems across manufacturing, logistics, and smart city infrastructure over the next decade.
