Considerations for Realizing Autonomous Manufacturing by AX
The expectation of autonomous manufacturing driven by AX (AI Transformation) is not limited to the domestic semiconductor manufacturing sector. Examining the VoCs of various manufacturers reveals that it represents an innovative shift in global manufacturing trends and a future-oriented trend. While high-end manufacturers are disclosing roadmaps for autonomous manufacturing to become first movers, the details remain confidential. Most manufacturers are currently at Levels 2 and 3 of the five levels of smart factories proposed by the KOSMO-Korea Smart Manufacturing Innovation Center (e.g., Level 1 - Inspection, Level 2 - Monitoring, Level 3 - Control, Level 4 - Optimization, and Level 5 - Autonomous Operation). For these reasons, we believe it is necessary to collectively consider the essential requirements for autonomous manufacturing within the ecosystem, and we aim to provide appropriate direction for the semiconductor manufacturing industry as a whole.
First, the lack of global standards or minimum guidelines for autonomous manufacturing is a significant challenge. Currently, even the SEMI standard (the standard for semiconductor manufacturing for the past 50 years) lacks a global standard, so discussions began in 2024 under the auspices of SEMI Korea, with the Semiconductor Manufacturing Ecosystem (SEM) gathering. The so-called Autonomous FAB Working Group was launched, with two sub-working groups: PM (Preventive Maintenance) Automation and Data Traceability.
Next, in addition to the conceptual definition of autonomous manufacturing, it is necessary to examine the realistic scope and regulations of automation. While the definition of autonomous manufacturing may vary depending on the specific realities and objectives of each manufacturer, it is often referred to in business circles as "autonomous manufacturing," "full automation," "unmanned automation," "lighthouse factory," "dark factory," and "AI autonomous operation." Therefore, this paper aims to distinguish between the meanings of "autonomous" and "automation," examine the differences, and propose a unique standard.
In conclusion, "automation" alone does not equal "autonomy." "Autonomous manufacturing" faces critical challenges, such as whether it can support the most rational decision-making even in exceptional and critical situations, propose solutions, and ensure process progression without interruption or with minimal disruption when problems arise. Finally, Miracom I&C proposes a realistic approach to autonomous manufacturing: an integrated platform (Miracom Digital Platform) based on SDF (Software Defined Factory) that supports AX decision-making, and smart equipment in manufacturing sites can self-diagnose to solve and prevent problems in advance.
