Laon People Drives Smart Factory Innovation with AI-Based ADC System

ADC System technology unveiled at the AI Machine Vision Conference

▲ In Se yong, Director at Laon People [Source: HelloT]

Laon People, a leading AI vision solutions provider, showcased its ADC (Automatic Defect Classification) System for Smart Factory implementation at the 2025 AI Machine Vision Conference.

“Incorporating AI, vision, and MLOps, the ADC System is a platform that analyzes manufacturing quality data in real time and connects the entire production line into a single network,” said Seyong In, Executive Director at Laon People. “It is a truly end-to-end solution for building smart factories.”

Laon People has long supplied automated systems across semiconductor, display, secondary battery, and mobility manufacturing, leveraging its advanced AI vision inspection technology. The newly unveiled ADC System goes beyond inspection equipment—it is an integrated MLOps-driven platform that manages AI model operation, validation, deployment, training, and feedback to maximize productivity and ensure stable quality in manufacturing environments.

AI Vision + MLOps Integration Enables Real-Time Quality Feedback

The ADC System combines Laon People’s flagship vision solution NAVI AI PRO, its MLOps platform, and industrial safety monitoring technology. Using vision data collected on-site, the AI model classifies defect types and delivers real-time quality feedback by linking with external systems such as MES, ERP, and NAS.

The system is also compatible with existing inspection machines and supports model retraining by loading stored inspection data from NAS. AI models automatically fine-tune quality thresholds based on confidence scores, optimizing defect rates and inspection speed across processes.

Field Deployment Demonstrates Fully Automated Learning–Validation–Deployment Cycle

Laon People demonstrated an end-to-end automated workflow based on real-world factory deployments. The ADC System manages defect and label data for various product families through a unified platform, automatically handling data collection, training, and model deployment.

In the demo, the number of review-required cases decreased from nine in the first inference to seven in the second inference after automated learning and redeployment—evidence of the system’s model improvement capabilities.

“By fully automating the review–training–redeployment cycle, we have significantly accelerated on-site response times and achieved standardized process efficiency and defect data management,” said In.

Confidence Score–Based Quality Control

The ADC System allows users to define quality targets for each product category and dynamically adjusts AI intervention levels based on configurable confidence score ranges (e.g., 87, 83, 78). Through the combined operation of AI Vision, MLOps, and Inspector modules, the system quickly adapts to changes in production environments, continuously recalibrating model performance to reduce defect rates while lowering manpower requirements.

“ADC System resolves the long-standing problem of fragmented process data and establishes a framework where AI autonomously makes quality decisions,” In noted. “AI vision is evolving into the central brain of the smart factory.”

Expansion Across the Manufacturing Value Chain

Laon People plans to extend the ADC System across the entire manufacturing value chain, including production, assembly, packaging, loading, warehousing, shipping, and safety operations.

“The essence of a smart factory is not mere automation but the transition toward a data-driven decision-making system,” said In. “The fusion of AI vision and MLOps will be the core technology enabling factories that continuously evolve.”

The company is currently developing customized ADC models tailored to key industries such as secondary batteries and high-precision assembly processes.