SIZL Demonstrates Manufacturing Humanoid on Secondary Battery Production Line, AI Guides the Process by Moving to Equipment
TECHWORLD ·
✦ AI Summary
SIZL conducted an on-site demonstration of its industrial humanoid robot, "Factory Copilot," on KGA's secondary battery electrode production line.
The robot understood workers' voice commands, moved autonomously, and provided production information linked to MES as well as environmental information such as temperature and humidity.
SIZL plans to expand the robot's functions to include equipment anomaly detection, root-cause analysis of quality issues, automated work instructions, and automated reporting.
AI-based smart manufacturing platform company SIZL unveiled an industrial humanoid robot that understands voice commands from factory-floor workers, moves autonomously, and provides production information. SIZL said it carried out an on-site demonstration of "Factory Copilot" on KGA's secondary battery electrode production line.
The demonstration took place in a real production environment where Coater and press equipment are operated. SIZL checked autonomous driving, voice interaction, process guidance, and environmental monitoring in this setting.
Factory Copilot connects natural-language instructions from workers to actual movement. For example, when a worker asks the robot to save the location of a new press machine, it recognizes surrounding objects and the environment with Vision technology, determines 3D spatial information with LiDAR sensors, and then avoids obstacles to move to the machine.
After arriving at the equipment, the robot calls up MES-linked information and provides voice guidance on work progress. The voice guidance items are production progress, production LOT, and estimated completion time. In the demonstration, the process of confirming that the equipment's progress was 80% was shown.
At the same time, the robot was equipped with RAG-based manufacturing knowledge linkage and KAG-based manufacturing knowledge linkage technologies. Using KGA electrode process standard operating procedures (SOP), equipment manuals, quality management standards, and failure-response databases, the robot provides workers with the information they need in response to questions.
When an on-site engineer asked about the current work process, the robot determined based on location information that it was the coating process. It then explained that the process applies slurry to a current collector and dries it, and also described the quality management factors for the coating process. It also guided the engineer through conditions that affect coating quality, including slurry condition, coating amount and thickness, line speed, and drying conditions.
On manufacturing sites, the ability to check the necessary information immediately in front of equipment is driving a field-centered expansion of AI-production data linkage methods. This demonstration included SIZL's factory data query function.
In this demonstration, SIZL added real-time environmental information analysis on top of equipment data queries, and confirmed the temperature around the Coater equipment at 23.4 degrees and the humidity around the Coater equipment at 70% through linkage with Wi-Fi-based IoT environmental sensors. It then demonstrated the process of comparing temperature and humidity with process management criteria.
In addition, when asked a question assuming that management criteria had been breached, SIZL provided SOP-based response measures, including calling the responsible engineer and checking the air conditioning system. Along with this, voice interaction optimized for the high noise levels of the production line was demonstrated, and industrial Dialogue Management technology was applied to handle worker follow-up questions, Barge-in, topic changes, and corrections.
As SIZL pushes ahead with developing software-based manufacturing AI agents, it is expanding the scope of its technology application by combining manufacturing AI agents with robotics into Physical AI. SIZL's Co-pilot AI won the grand prize from the Ministry of SMEs and Startups in 2025.
Lee Ji-hyun, CEO of SIZL, said existing industrial robots were limited to preprogrammed routes and repetitive tasks. She added that the company's SIZL humanoid concept links company data and Copilot AI, allows the robot to autonomously perceive the site, communicate with workers via voice, and move and operate as an intelligent object.
SIZL conducted an on-site demonstration at KGA, and Lee said this confirmed the robot's ability to converse in a production setting, understand complex site information, and connect site information with autonomous robot behavior. SIZL has accumulated smart factory and manufacturing AI technologies, and plans to develop them further into a Physical AI platform specialized for global manufacturing sites by combining accumulated technologies with hardware robotics.
SIZL plans to expand Factory Copilot's functions in the future to include autonomous on-site patrols, equipment anomaly detection, root-cause analysis of quality issues, automated work instructions, and automated reporting. In parallel, SIZL is preparing for a KOSDAQ AI technology special listing (IPO) targeted for 2027, while pursuing technology transfers with a research institute under the Ministry of Trade, Industry and Energy and also promoting national projects related to smart manufacturing.
Source: TECHWORLD · Kim Gyeong-ju
Original: https://www.epnc.co.kr/news/articleView.html?idxno=407402
References
This article was produced with the help of an automated content generation algorithm.
Source: TECHWORLD
View originalThis article was summarized and organized by BizCrush based on the original article from TECHWORLD. For exact quotations and full details, please refer to the original article.