[Edaily Reporter Kim Hyun-ah ] SeoYeon E-Hwa ( SEOYONEHWA(200880)), a company specializing in automotive interior and exterior parts, is set to fully launch its “Manufacturing AX (AI Transformation)” initiative, which aims to transition the manufacturing experience and on-site knowledge accumulated over 50 years into an artificial intelligence (AI)-based operational system.

SEOYONEHWA announced on the 3rd that it has selected the AI software company KONAN TECHNOLOGY INCORPORATION(402030)as its partner for building manufacturing Agent AI agents and will pursue an AI innovation project centered on the manufacturing floor in collaboration with JisanSoft.

The goal of this project is to go beyond simple task automation by applying AI agents to the entire manufacturing process—including design, tooling, and quality control—to accelerate decision-making and boost productivity.

Founded in 1972, SEOYONEHWA is a leading South Korean automotive parts manufacturer that produces interior and exterior components such as door trims, seats, and bumpers. The company has grown by supplying parts to domestic and international automakers, including the HyundaiMotor Group, and has established itself as a global automotive parts company with annual consolidated revenue exceeding 4 trillion won.

In particular, the company has secured a competitive edge in the fields of design, tooling, and quality control based on the manufacturing experience and process know-how it has accumulated since the production of interior parts for HyundaiMotor’s Pony model.

The core of this AX project is to transform the “tacit knowledge” accumulated within SEOYONEHWA into knowledge assets that generative AI can utilize.

KONAN TECHNOLOGY INCORPORATION will develop 20 specialized Agent AI systems to be applied to SEOYONEHWA’s core business systems, including △S-ENG (process design), △S-MLD (mold development), and △S-QCX (quality control).

The plan is to digitize manufacturing know-how—which previously relied on the experience and judgment of skilled workers—so that AI can learn and utilize it, thereby helping to prevent design errors, improve quality, and reduce problem-solving time.

In particular, this project will utilize “multi-agent orchestration” technology. This approach goes beyond a single AI providing simple responses; instead, multiple specialized Agent AI agents collaborate to set goals, formulate work plans, execute tasks, and verify results.

For example, if a quality issue arises, AI agents that analyze relevant data, identify past similar cases, and propose improvement measures will work together to support on-site decision-making.

JisanSoft will be responsible for the overall AX implementation and development management based on its low-code development platform “LAMP7,” while KONAN TECHNOLOGY INCORPORATION will handle the development of manufacturing-specific AI solutions and the agent system.

The two companies are also pursuing the development of a joint solution that enables AI to utilize unstructured document knowledge from manufacturing firms. The plan is to analyze Excel and PowerPoint files, as well as drawings containing past vehicle problem cases, root cause analyses, and corrective actions using Vision Language Models (VLM) and Optical Character Recognition (OCR) technology to create a knowledge database.

This will enable on-site employees to quickly search for similar past cases and apply accumulated manufacturing experience to their work.

KONAN TECHNOLOGY INCORPORATION plans to use this project as an opportunity to accelerate the expansion of its “Physical AI” business, which operates directly on the manufacturing floor. The company is currently collaborating with domestic universities to research smart manufacturing technologies based on Visual-Language-Action (VLA) AI—which integrates vision, language, and behavior—and Large-Scale World Models (LWM) that understand and predict the real world. It also plans to utilize this research to further enhance “VisionFlow,” a wearable remote-support Agent AI designed for the manufacturing environment.

Kim Young-seom, CEO of KONAN TECHNOLOGY INCORPORATION, stated, “The SEOYONEHWA AX Project is a case study demonstrating how generative AI and Agent AI can transform the competitiveness of the traditional manufacturing industry,” adding, “We will support the AI transformation of manufacturing companies based on domain-specific AI technology.”