Hanwool Semiconductor Targets Global Expansion Market with AI Full Inspection: "13,000 MLCCs Inspected Per Minute"

Growing Demand for High-Reliability MLCCs for AI Servers and Electric Vehicles
Comprehensive Six-Camera Imaging of Products
Combining Rule-Based Image Processing and AI Judgment
Expanding Supply Scope from External Inspection to Multifunction Measurement, Ultrasonic Internal Inspection, and Automation Equipment

As the adoption of artificial intelligence (AI) servers and electric vehicles accelerates, both the performance and reliability standards for multilayer ceramic capacitors (MLCCs) have been raised. As components become smaller and the number of internal layers increases, even minute defects can impact the reliability of products. As a result, AI-powered full inspection equipment is emerging as a new source of competitive advantage in the manufacturing process.

According to the securities registration statement submitted by Hanwool Semiconductor on July 16, the company has outlined a strategy to target demand from global MLCC manufacturers seeking to expand production capacity or replace outdated equipment, focusing on its AI-powered MLCC six-sided external inspection equipment.

MLCCs are essential passive components used in nearly all electronic devices, including smartphones, home appliances, automobiles, servers, and industrial equipment. Recently, the rapid expansion of markets such as AI servers, electric vehicles, autonomous vehicles, and industrial robots has driven a surge in demand for high-capacity, high-voltage, and ultra-compact MLCCs.

In particular, MLCCs for AI servers and automotive electronics must operate stably under high-temperature and high-voltage conditions. Even small external cracks or internal defects can affect performance and lifespan, so much stricter quality control is required compared to consumer-grade products.

To address these evolving market needs, Hanwool Semiconductor has made its AI-powered inspection equipment a flagship product. The equipment operates by using a vibration supply unit and the centrifugal force of a high-speed rotating glass disc to transport ultra-compact MLCCs in a continuous flow. As the products move, six high-resolution cameras capture images of all sides, which are then analyzed in real time by the company’s proprietary AI inspection platform, "HawAIe," and AI analysis engine, "AlohaNet."

According to the company, this equipment can process approximately 13,000 pieces per minute when inspecting 0603 size MLCCs. It checks for a range of external defects, including breakages and chipping, surface cracks, foreign substances, and electrode shape irregularities.

The 0603 MLCC refers to an ultra-compact component measuring about 0.6mm by 0.3mm. Given that these are similar in size to a human hair, the equipment must transport products at high speed, perform simultaneous imaging from six directions, and determine defects in real time, making both mechanical precision and image analysis technology essential.

Hanwool Semiconductor notes that the combination of traditional rule-based image processing techniques and AI judgment technology is a key competitive edge. Defects with consistent location and shape are determined by preset criteria, while non-standard cracks and micro surface flaws that are difficult to identify using conventional methods are analyzed by AI.

This approach allows them to maintain inspection speed while minimizing false positives—normal products being judged defective—as well as false negatives—actual defects that go undetected. The company also plans to further enhance accuracy by leveraging accumulated inspection data for AI learning, accommodating the wide range of defects that vary according to product specifications and production processes.

External defects in MLCCs do not occur in a single form. Types include chipping (where product corners break), surface cracks, structural damage, electrode shape irregularities, and surface contamination. Even within the same type of crack, the location, length, width, and contrast can vary significantly.

Additionally, tiny marks formed during production may appear similar to normal surface patterns or light reflections, making accurate judgments difficult with simple preset criteria. To address these variations, Hanwool Semiconductor designed its equipment to capture all sides of the product using six cameras and to employ both rule-based image processing and AI analysis to identify various types of defects in a single process.

Starting with external inspection equipment, the company is now expanding its product lineup to encompass a full range of MLCC post-processing equipment. Hanwool Semiconductor has successively developed multifunction measurement instruments, automatic mounting machines, ultrasonic non-destructive inspection equipment, cutters, transfer machines, and process automation facilities to build a comprehensive post-processing lineup.

The multifunction measurement instruments test electrical characteristics such as capacitance, insulation resistance, and withstand voltage. Ultrasonic non-destructive inspection equipment detects internal cracks, delamination, and voids that cannot be identified externally. The automatic mounting machine precisely positions ultra-compact MLCCs, while cutters and transfer machines handle slicing and moving of stacked structures for subsequent processes.

Hanwool Semiconductor is pursuing a business model that extends from external inspection equipment to measurement, internal inspection, and automation equipment after gaining entry to customers' production lines. This strategy aims to develop the company beyond a mere supplier of individual equipment into a provider of solutions that cover the entire MLCC post-processing segment.

However, expansions in production facilities by global MLCC manufacturers do not immediately translate into equipment orders. To win new customers, the company must complete several steps, including equipment performance evaluations, sample testing, validation for mass production line application, and negotiations over price and delivery schedules. Ultimately, the key factors for actual business success will be whether the AI-based inspection technology and processing speed satisfy mass production requirements and how many new overseas customers can be secured.

A Hanwool Semiconductor representative commented, "With AI servers and electric vehicle MLCCs, even the most minute defects can affect product reliability. As production capacity increases, the importance of full inspection technology is also on the rise. By combining high-speed transfer technology, six-sided imaging, and AI defect assessment, we are enhancing both the speed and accuracy of inspections."

The official added, "We plan to use external inspection equipment as a springboard to expand our supply scope into multifunction measurement, automatic mounting, ultrasonic internal inspection, and process automation equipment, ultimately building a product portfolio that covers the full range of post-processing equipment for MLCCs."

This content was produced with the assistance of AI translation services.