This video demonstrates the live operation of an AI-powered SOP Assembly Behavior Analysis System for intelligent quality control on a lighting product assembly line.

Industrial cameras continuously monitor operators performing standardized assembly tasks at a workstation equipped with a green fixture table. Multiple black rotor components are positioned along the conveyor line, while assembly tools—including an adhesive dispensing bottle, electric screwdriver, torque socket wrench, metal pipe fittings, and screw trays—are arranged within the operator's working area.

The operator, wearing a blue work vest, gray long-sleeve uniform, and white protective gloves, is performing a screw-tightening operation using a torque socket wrench. The AI vision system automatically tracks the operator's hands and tools, detects hand keypoints in real time, and accurately recognizes the tightening action through advanced human pose estimation and operator behavior recognition algorithms.

The system verifies every assembly step according to the predefined Standard Operating Procedure (SOP), including retaining ring installation, screw fastening, rotor placement, nut installation, adhesive dispensing, and rotor block assembly. All inspection items for the current product have successfully passed the SOP compliance check and are marked as PASS.

The intelligent dashboard continuously visualizes key production metrics, including:

• Individual process cycle time (PT)
• Total production cycle time (CT)
• Historical OK and NG counts
• Defect rate
• Yield rate
• Total inspected products

The system distinguishes between two categories of defects:

• Operator Defects – Non-standard or incorrect assembly actions.
• Product Defects – Quality issues caused by defective workpieces.

To support complete production traceability, the system automatically captures thumbnail images for every assembly step and stores synchronized video recordings, allowing engineers to review historical operations and investigate quality issues.

The platform provides comprehensive AI-based SOP compliance monitoring, automatically determining whether assembly operations conform to standardized procedures while collecting production time data, tracking operator performance, analyzing quality statistics, and enabling digital management of assembly behavior throughout the manufacturing process.

At the time of recording, the system is operating in continuous production monitoring mode and is inspecting the fourth product on the assembly line. All assembly operations have been completed successfully, with no SOP violations or abnormal operator behaviors detected.