person holding tool during daytime

SolVisionCase Study

Automotive

Visual Inspection of Automotive Weld Beads

Detecting defects in random brightness conditions

Welding of car body and structural quality 

The structure of car bodies are complicated and made up of multiple stamped metal parts connected usually by welding, bonding, or riveting. Modern cars often use low-carbon steel plates, so welding is a common technique in the manufacturing process. During welding, the weld pool absorbs gas from internal reactions or external dissolution. Trapped gas sometimes cannot be discharged in time and remain in the weld bead, forming pores that affect the appearance and also structural quality of car bodies. 

Detecting defects in random brightness conditions

Brightness levels and images of weld bead porosity

Automobile roofs come in shapes that require weld beads in different heights, and so brightness conditions can vary when they are pictured. Randomly generated gas pores also have inconsistent appearances and characteristics, making it difficult to define rules for weld bead defect inspections.

Overcoming inconsistent brightness with SolVision

Powered by AI, SolVision can facilitate simulations of potential weld bead defects in different brightness to train the AI model, so that inspections of car bodies for weld defects are reliable and not constrained by brightness or lighting conditions. 

AI Inspection

Golden Sample

Car roof weld bead detection case

Defective Weld Beads

Car roof weld bead detection case

Related Posts

  • Automated Recognition of License Plates & Cargo Container Numbers

    Solvison enables OCR for license plates and container numbers even when shapes, color and locations of characters vary, and are pictured in poor lighting conditions.
  • a group of square objects

    Detecting Chipping Defects in Wafer Dicing

    The location, size and shape of cracks vary each time, and traditional optical inspection cannot accurately identify such unpredictable defects.

  • Inspecting Packaged Semiconductor Chips

    The Anomaly Detection tool uses deep learning technology to teach the AI model sample images of “perfect” wafers.
  • a close-up of a machine

    AI Laser Welding Defect Detection

    As welding positions and patterns are ever-changing, the number of possible variations makes it challenging for AOI systems to recognize welding weave patterns or locate missed welding spots and other production errors.