VisionPro Deep Learning
AI-powered PC image inspection software for fast, robust quality control.
The Challenge: Mission-Critical Components with Zero Defects
Heat shield components are made from high-cost, specialized materials, such as thin woven polymer composites or thick white foam blocks. During the manufacturing process, these materials are prone to defects such as:
- Creases: imperfections from handling and processing
- Tears and cuts: damage that can propagate under stress
- Gouges and discoloration: surface defects that can affect thermal protection
Manually inspecting these materials for subtle defects is slow, inconsistent, and costly. Missing a single defect can mean scrapping an entire production run. Traditional machine vision systems and rule-based solutions cannot account for defect variability or adapt to new defect types.
The Solution: Powerful, User-Friendly Machine Vision
Cognex developed AI that helps machine vision systems reliably identify and classify defects. Vision systems use robust image-formation technology to maximize contrast on complex surfaces, creating clear, detailed images. Then, AI-powered software analyzes images of defective and acceptable heat-shield materials, enabling the vision system to detect subtle defects. Since the solution learns by comparing images of acceptable and defective components, it can recognize new and variable defects.
Automating heat shield inspection provides:
- Consistency: eliminate the subjectivity of manual inspection with uniform, rigorous standards
- Cost reduction: lower scrap rates by detecting defects earlier and prevent false rejects
- Scalability: manufacturing and imaging data help build a foundation for new projects