AI-based 360-degree Visual Inspection System for Die-cast Products
About IVI-360™
IVI-360™ is an AI-powered automatic visual inspection system developed for aluminum die-cast and cast products.
Born from real-world challenges encountered in die-casting production sites, this system was designed to be easy to set up and highly durable, ensuring smooth operation even in mass-production environments. Its user-friendly configuration makes it suitable for small to medium-sized factories as well as overseas plants.
With seven degrees of freedom (a six-axis robot plus a rotary stage), the system automatically captures images of the inspection target from all 360 degrees, while the AI automatically detects and classifies defects.
The name “IVI-360™” concisely expresses the system’s technical concept and core features. “IVI” stands for Intelligent Vision Inspection, representing AI-based image analysis. The number “360” symbolizes the system’s key advantage—its ability to capture the entire surface of a product from every angle using a single camera and rotary stage, without the need to invert the workpiece.
“NACHi” is a registered trademark of NACHI-FUJIKOSHI CORP. (JPO Registration No. 4340845). “IVI-360” is a registered trademark of The 7th Engineering Co., Ltd. (JPO Registration No. 6941983).
The Role and Challenges of Human Visual Inspection
In developing IVI-360™, we placed great importance on two guiding principles: grounding our design in the realities of die-casting production sites and taking practical steps toward achievable automation. Our goal is not to simply replace human inspectors with automation, but to respect the essential human strengths of judgment and flexibility while reducing workload and ultimately stabilizing product quality.
In manufacturing fields such as agricultural machinery and automotive components, skilled inspectors still play a crucial role in identifying subtle defects based on experience. Their delicate sense of judgment cannot be easily replicated by AI or vision systems alone.
Especially in die-casting, minor changes in molds or process conditions naturally cause some variation in the products. Maintaining stable quality under such variability requires flexible decision-making at the inspection stage.
Of course, the foundation of quality control lies in “preventing defects at the source.” Measures at the upstream stages—design, mold control, and casting condition management—are the most important. However, in reality, constraints such as cost and delivery schedules make it difficult to address all issues upstream. For this reason, inspection still plays a vital role in preventing defective products from being shipped.
Against this backdrop, IVI-360™ has been designed as a partner that supports quality management professionals. In the demanding environment of die-cast visual inspection—where focus and attentiveness are constantly tested—we aim to ease the burden on operators and contribute to the creation of stable, consistent product quality together.
Advantages and Disadvantages of Visual Inspection
⭕️ Advantages of Visual Inspection
☺️ Flexible Judgment: Can immediately adapt to variations in shape or changes in inspection criteria.
☺️ Low Initial Cost: No need for dedicated equipment, allowing quick setup.
☺️ Expert Judgment Based on Experience: Experienced inspectors can detect subtle defects that machines may overlook.
☺️ Suitable for Small-Lot, Multi-Product Production: Easily adaptable to environments with frequent product changes.
❌ Disadvantages of Visual Inspection
😢 Variation in Accuracy: Standards and sensitivity differ between inspectors.
😢 Mistakes Due to Fatigue or Loss of Focus: Long working hours can lead to oversight and misjudgment.
😢 Need for Training and Maintenance: Continuous education and supervision are required to maintain consistent standards.
😢 Weak in Traceability: Inspection results are often not well-documented or standardized.
While visual inspection is flexible and adaptable to diverse situations, it is also highly dependent on individuals and subject to variability and errors caused by fatigue or mood.
In modern manufacturing, a hybrid inspection approach—combining human visual inspection with AI or machine-based inspection to reduce fatigue—is becoming the mainstream practice.
What Is AI Anomaly Detection?
AI anomaly detection is a technology that enables AI to automatically identify “unusual conditions or patterns.”
For example, in die-casting operations, it is essential to quickly and consistently detect issues such as poor metal flow, burr residue, or impact marks.
Traditional inspection methods are rule-based, relying on fixed criteria such as thresholds or dimensions to detect anomalies. While effective for simple defects, these methods often struggle with complex shapes or subtle irregularities.
AI, on the other hand, learns from large volumes of image data—both of “normal” and “defective” products—and gradually develops its own understanding of the differences. As a result, it can detect defects with a level of accuracy comparable to human visual inspection.
The illustration depicts a scene where a single bird is hidden among many leaves—symbolizing how living beings instinctively notice something that “doesn’t belong.” Humans don’t just see shapes or colors; we perceive imbalance and inconsistency within context.
By mimicking this capability, AI becomes better at recognizing anomalies more accurately and efficiently.