AIVEX Implements AI Vision Robotics-Based Autonomous Automation System for a Leading Secondary Battery Materials Company

May 29, 2026, Seoul – AIVEX, an industrial Physical AI company, announced that it has successfully deployed an unmanned automation solution for the crucible packaging dismantling process at a leading South Korean secondary battery materials company, based on its AI robotics control platform ‘AIVot’.

The project focused on automating the removal of wrapping film and straps from crucible pallets used in the cathode material production process, a task that had previously been performed entirely manually by operators. As a result, despite the production line operating 24/7, nighttime and weekend operations were limited, while safety risks for workers remained a concern.

Due to the varying number, position, and orientation of straps on each crucible, conventional fixed automation approaches were unable to handle the process. In particular, vertical straps were difficult to detect reliably with conventional optical systems due to visibility variations caused by lighting reflections and changes in material angles. In addition, areas where straps intersect required accurate classification of horizontal and vertical strap positions, demanding highly precise AI vision capabilities.

AIVot is an AI robotics platform developed based on ROS2, integrating 2D and 3D optical systems (hand-eye type), NVIDIA Jetson PC, AI models, 6D Pose Estimation, automated robot path generation, and communication with peripheral equipment. Based on AIVot, AIVEX developed and deployed a customized robotic automation solution, enabling rapid development of an adaptive AI robotics solution capable of flexibly responding to environmental changes and variations in workpieces.

Within the process, AIVot performs the following tasks:
△ Strap detection and automated path generation
△ Strap joint detection
△ Automatic determination of vertical strap direction (top/bottom orientation)
△ Band position detection
△ Wrapping film cutting path generation based on crucible position correction
△ Post-cutting inspection

In particular, to handle varying working conditions for each crucible, the system was designed to generate a new path plan for every cycle, enabling flexible operation across diverse working environments without additional robot teaching.

Through this deployment, the customer achieved:
△ 100% unmanned automation of the crucible packaging dismantling process
△ Establishment of a 24/7 continuous operation system
△ Zero worker safety incidents
△ Achievement of a tact time within 14 minutes

AIVEX CEO Minsoo Sung said, “Automation in industrial environments is evolving beyond simply performing repetitive tasks mechanically. The next generation of automation enables AI to understand and interpret environments, make decisions, and connect those decisions directly to robotic actions. This project is meaningful as an example of Physical AI-based manufacturing automation, where AI perceives real-world conditions and determines robot motion planning.”

For this project, AIVEX provided the overall vision solution, including optical systems, lighting, vision controllers, and AI algorithms. The partner company was responsible for robot hardware, grippers, and overall equipment project management, while AIVEX focused on vision recognition, AI-based decision-making, and automated robot path generation.

Meanwhile, AIVEX plans to continue expanding the application of its Physical AI platform across various manufacturing environments based on its capabilities in in-house optical system development, AI software integration, and hybrid algorithm operation for handling variable environments.