SolVisionCase Study

Petrochemicals | Textiles and Footwear

AI Inspection of Plastic Buckles

Detecting defects on injection molded plastics

Quality of injection molded plastics 

Plastic buckles come in many forms and are mostly manufactured in large quantities by injection molding. Many variables can affect the outcome of plastic injection molding processes, including raw materials, molds, temperature, pressure, and production speed.

Surface defects and traditional inspection systems

Production flaws commonly seen on plastic buckles range from oil stains and white spots, to burrs and debris. Amongst them oil stains are the most difficult to be detected manually or by  rule-based systems, as they are generally less noticeable.

Multi-tool AI defect detection

Powered by artificial intelligence, SolVision features four individual tools designed for different inspection tasks. To identify multiple defects at the same time, the system’s Segmentation and Classification tools utilize deep learning technology to perform recognition of faulty products, and train an AI model capable of detecting defects accurately, including oil stains, in real time.

AI Inspection

Oil Stain

Plastic buckle defect detection case

White Stain

Plastic buckle defect detection case

Burrs

Plastic buckle defect detection case

Burrs

Plastic buckle defect detection case
Related Posts
  • assorted coffee capsules

    Product Packaging Sequence Detection Using AI

    AI-driven precision in coffee capsule packaging, overcoming color and reflective metal challenges for automated packaging sequence detection.
  • robot arms controlled by VGR operating on an automotive production line

    Vision Guided Robotics in Automotive Processes

    Solomon’s VGR solution, Solmotion, uses cutting-edge AI and 3D technologies to assist complex manufacturing operations such as car sealing.
  • food quality inspection of apples on a conveyor in a factory

    Food Quality Inspection Using AR + AI

    META-aivi enables real-time detection of foreign objects in food production, addressing challenges associated with manual inspection such as worker fatigue and the risk of human error.

  • Detecting Adhesion Defects in Semiconductors

    Excessive adhesive may remain on the chip or overflow on the circuit board and cause the chip to tilt, affecting stability of the whole semiconductor package.