The platform manages the full cycle of robot deployment, from data collection and training to operation
Forge trains robots for industrial sites, while Baton controls different types of robots through one system
Hyun Shin-gyoon, president of LG CNS, delivers welcoming remarks at RX Media Day held at LG Science Park in Magok, Seoul, on the morning of May 7.
[Insight Korea = Nam Bit-haneul]
“The core of robot transformation, or RX, is not the performance of individual robots. It lies in building a system for learning, verification and integrated operations tailored to each work site. LG CNS aims to present its answer to RX through the PhysicalWorks platform.”
Hyun Shin-gyoon, president of LG CNS, made the remarks on the morning of May 7 at the company’s RX Media Day held at LG Science Park in Magok, Seoul, where he unveiled “PhysicalWorks.” PhysicalWorks is LG CNS’s proprietary RX platform brand, designed to manage the entire life cycle of robot adoption — from data collection to training, verification, on-site deployment, operation and monitoring — in one integrated system.
As global companies accelerate competition to commercialize industrial robots, LG CNS is focusing less on robot hardware itself and more on enabling robots to understand real industrial work and perform it reliably. The company sees robots evolving beyond simple automation equipment into active agents that carry out production and operations.
“We view the entire process as one system: going beyond robot adoption, training and verifying robots so they can actually perform tasks on-site, operating them stably after deployment and continuously upgrading them,” Mr. Hyun said. “I believe that is the essence of RX, which innovates productivity and operating methods in industrial settings.”
Forge Supports the Full Robot Life Cycle
LG CNS’s PhysicalWorks consists of two platforms: PhysicalWorks Forge, or Forge, which supports robot training, and PhysicalWorks Baton, or Baton, which manages different types of robots in an integrated way.
Forge, whose name suggests the idea of training or shaping, was developed to train robots to a level at which they can be deployed in real industrial settings. Its defining feature is that the full process, from collecting data needed for robot learning to verification and site application, can be handled in one place.
Its biggest point of differentiation is how it secures data. In the past, robots had to imitate human behavior thousands of times. With Forge, robots learn from data in a 3D virtual environment that recreates actual industrial sites and tasks. LG CNS plans to add methods such as converting videos of human work into training data and using motion capture.
Forge also automates the process of selecting acquired data. AI automatically selects, organizes and processes data useful for robot learning. For example, when a robot learns to pick up products in a factory, the system selects only successfully completed motions and removes unnecessary segments to improve efficiency.
After training, robots undergo simulation in a 3D virtual environment to verify their ability to perform tasks and operate safely. They are then optimized for the actual site and deployed. LG CNS says this can shorten the time from robot training to field deployment from several months to one or two months.
“With Forge, customers can secure the data needed for training in the fastest way and train robots easily even without experts in robot brains,” said Park Sang-yeop, chief technology officer at LG CNS. “Forge can provide this practical value to customers.”
Baton Integrates and Controls Different Robots
Baton is the platform that assigns tasks to trained and deployed robots and provides integrated control and monitoring. As its name suggests, Baton acts like a conductor’s baton, allowing robots from different manufacturers to be operated within one system. It addresses the problem that robot management becomes more complex as the number of robot types grows, because each manufacturer uses different control methods and operating interfaces.
According to LG CNS, applying Baton to an environment with about 100 robots, including autonomous mobile robots, or AMRs, and automated guided vehicles, or AGVs, is expected to raise productivity by about 15 percent and reduce operating costs by up to 18 percent.
“For robots to work in the field, more than 120 technologies, more than 50 experts and over 10,000 hours are needed,” Mr. Park said. “Through the PhysicalWorks platform, these things can be done the fastest, with the highest quality and at the lowest cost.”
The effectiveness of PhysicalWorks was demonstrated at the event. For the first time in Korea, different types of robots were shown autonomously collaborating without human control. Four types of robots trained through Forge — bipedal, quadruped, wheeled and AMR robots — worked together in a logistics setting based on Baton.
In the demonstration, a bipedal robot placed items into a box and loaded the box onto a quadruped robot. The quadruped robot then carried it to a wheeled robot. The wheeled robot received the box and placed it on a shelf, with the entire process unfolding organically.
The two platforms are already being applied in real-world settings. According to LG CNS, Forge is currently being used in robot proof-of-concept projects with more than 20 customer companies. Baton is being used in the Busan Smart City National Pilot City project to integrate and monitor four types of robots, including patrol, barista, luggage-carrying and cleaning robots.
LG CNS said it aims to make the business results of PhysicalWorks visible within two years. Lee Jun-ho, executive vice president and head of LG CNS’s Smart Logistics & City business unit, said, “We will first apply it to logistics and manufacturing sites, and based on that, accumulate and expand industry-specific robot foundation models, or FRMs.”
“It will take about two years for the results to become visible in earnest,” he added.
Mr. Hyun concluded by saying, “Based on our deep understanding of industrial sites and our capabilities in robot automation and AI system integration, LG CNS aims to become a key partner that accompanies customers through the entire RX process.”
“In an era when robots become a source of industrial competitiveness,” he said, “LG CNS will open that path first.”
-This article was translated by ChatGPT. (이 기사는 챗GPT가 번역했습니다.)
댓글
0