Our company offers a comprehensive solution portfolio in seafood processing automation, intelligent smart factories, and digital healthcare, built on AI vision and Physical AI technologies. Our key products include an AI Vision-Based Automated Skipjack Tuna Head-Cutting System, which recognizes the shape, size, and orientation of each fish, determines the optimal cutting position, and interfaces with cutting equipment, as well as an AI Vision-Based Skipjack Tuna Byproduct Sorting System, which detects and classifies byproducts such as milt and roe and integrates them with robotic sorting processes. In the industrial sector, we are developing an MES–HACCP Integrated AI-Based Smart Factory Solution that collects and analyzes MES, HACCP, equipment, and inspection data in real time, detects early signs of process abnormalities, and centrally manages inspection, decision-making, corrective action, and record histories. By integrating AI vision, deep learning-based object recognition, data analytics, PLC and robotic control, and integrated monitoring technologies, the solution supports production automation, consistent product quality, and improved operational efficiency. In the medical and healthcare sector, our solutions include a Closed-Loop Mattress and Safety Guard System for Fall and Pressure Injury Prevention, which combines RGB-D-based non-contact posture recognition with a pneumatic mattress and protective guard; an AI-Based Health Monitoring System Using 2D Nail Images, which analyzes changes in nail color and shape to monitor potential health abnormalities; a Multi-Colorimetric Film-Based Wound Microenvironment Monitoring Platform, which analyzes colorimetric responses caused by changes in pH and temperature; and an AI-Integrated Sleep Bioacoustic Diagnostic Platform that utilizes acoustic data. We provide intelligent solutions that improve productivity, quality, safety, and management efficiency across industrial and healthcare environments by integrating the entire process, from data collection and processing, AI model development, image and sensor analysis, and real-time monitoring to robot and actuator control, platform development, and on-site demonstration.

This seafood-processing automation solution uses camera imagery to recognize, in real time, the size, shape, orientation, curvature, and head–body boundary of each skipjack tuna, which is a naturally irregular raw material. An AI algorithm then determines the optimal cutting position for each individual fish and transmits the result to the cutting equipment. Unlike conventional systems that cut at a fixed position, this system analyzes differences in body shape and orientation for every incoming fish, thereby improving adaptability to actual processing conditions and ensuring more consistent cutting quality. The system operates through the following process: raw-material feeding and conveying, vision imaging, AI-based shape and position recognition, calculation of cutting coordinates, cutting-equipment control, result verification, and data storage. RGB and RGB-D cameras, together with industrial lighting, are used to collect information on the external shape and position of each fish. Deep learning-based object recognition technology then distinguishes the head, body, and boundary regions. The cutting coordinates calculated by the AI are transmitted to the PLC and drive system, while coordinate calibration between the camera and the cutting equipment converts the detected position into an executable cutting point. Data reflecting actual seafood-processing conditions—including changes in illumination, moisture, surface reflection, differences in fish size, and variations in orientation—are continuously incorporated to improve recognition performance and on-site applicability. Images, recognition results, cutting positions, and operating records generated during the cutting process are stored in a database and used for process monitoring and further AI model improvement. Through an integrated monitoring interface, operators can review raw-material recognition status, cutting coordinates, equipment operating conditions, and the occurrence of abnormalities. Accumulated process data can also be analyzed to identify cutting errors and recurring causes of abnormal operation. As a result, the system goes beyond basic equipment automation and enables data-driven quality management and process optimization. The system is designed to address key limitations of conventional manual head-cutting operations, including variations in quality caused by differences in worker skill, fatigue and safety risks resulting from repetitive work, and difficulties in securing skilled labor. By automatically determining the appropriate cutting position for each fish, the system reduces unnecessary loss of usable meat, improves cutting consistency, and increases both worker safety and production efficiency by automating repetitive and hazardous tasks. The continuous collection of standardized process data also supports production history management, quality traceability, and maintenance planning. Our company carries out the entire system-development process, including the collection, processing, and labeling of skipjack tuna image data; AI model development; camera and lighting system configuration; cutting-coordinate calculation; PLC and equipment integration; integrated monitoring; and on-site demonstration. Through validation in actual seafood-processing environments, we plan to secure technical reliability and operational stability. Beginning with skipjack tuna head-cutting automation, we will expand the technology to preprocessing, sorting, and byproduct classification processes for a wider range of fish species, thereby supporting the intelligent automation of the seafood-processing industry.