When a major U.S. lens manufacturer needed to eliminate inconsistency in their quality inspection process, they turned to Intelgic to design and build an intelligent, standalone inspection system capable of detecting the smallest surface and coating defects — reliably, repeatedly, and at production speed.
The Challenge
Precision optical lenses demand near-zero defect tolerance. Even minor surface irregularities — imperceptible to the untrained eye — can compromise optical performance, trigger product returns, and damage brand reputation in a fiercely quality-conscious market.
The manufacturer's existing inspection process relied heavily on manual visual examination by trained operators under controlled lighting. While skilled inspectors were effective, the process suffered from fundamental limitations that scale and consistency could not overcome.
- Inspector fatigue causing missed defects during long shifts
- Subjective judgment varying between inspectors and shifts
- No consistent documentation of defect type, location, or severity
- Inability to scale inspection throughput with production volume
- Multiple lens diameters requiring frequent fixture changeovers
- Sub-surface and coating defects nearly invisible under standard lighting
- Automated detection of surface scratches, chips, and coating defects
- Support for multiple lens diameters without workflow disruption
- High-resolution imaging under multiple lighting conditions
- Real-time AI analysis with pass/fail output per lens
- Defect classification and location mapping for traceability
- Standalone, operator-loadable system requiring minimal training
"The challenge with optical inspection isn't seeing defects — it's seeing every defect, every time, on every lens, regardless of who is working or how long the shift has run."
The Solution: Standalone AI Inspection Station
Intelgic designed and built a self-contained AI inspection station from the ground up — purpose-engineered for lens inspection with every hardware and software decision made specifically for optical surface quality analysis.
Custom Multi-Diameter Fixture System
One of the first engineering challenges was accommodating the client's range of lens diameters. Intelgic developed a set of custom-precision lens fixtures — each machined to hold specific diameter ranges securely and repeatably in the exact same position relative to the imaging system. Operators simply select the appropriate fixture, load the lens, and initiate inspection. No alignment jigs. No guesswork. Repeatable positioning is guaranteed by the fixture design.
High-Resolution Multi-Condition Imaging
The imaging subsystem is the heart of the inspection station. Intelgic integrated a high-resolution industrial camera with a custom optical configuration designed specifically to reveal lens surface characteristics at the required detection threshold.
Critically, the system captures multiple images of each lens under carefully programmed lighting conditions — directional dark-field illumination to reveal surface scratches, diffuse brightfield to assess coating uniformity, angled ring lighting to surface edge chips, and UV or polarized lighting to identify sub-surface anomalies and coating delamination that would be invisible under conventional inspection light.
Operator Loads Lens
Operator selects the correct diameter fixture, places the lens, and initiates the cycle via a single button press. The system handles everything from this point forward.
Multi-Condition Image Capture
The imaging system sequences through multiple lighting configurations — dark-field, brightfield, ring, and specialized lighting — capturing high-resolution images under each condition in under 4 seconds.
Intelgic AI Analyzes Each Image
Intelgic's trained vision AI models analyze each image independently, cross-referencing findings across lighting conditions to distinguish genuine defects from reflections, condensation, or imaging artifacts.
Defect Classification & Mapping
Detected anomalies are classified by type, measured for size, and mapped to their location on the lens surface — generating a complete defect record for each inspected unit.
Pass / Fail Output & Traceability
The system issues a clear pass/fail decision based on configurable acceptance criteria, displays annotated defect images for operator review, and logs the full inspection record with timestamp and defect data.
Defect Types Detected
The system is trained and validated to reliably detect the full spectrum of defects relevant to precision optical lens manufacturing:
Surface Scratches
Edge Chips & Fractures
Coating Voids & Pinholes
Coating Delamination
Inclusions & Bubbles
Cosmetic Blemishes
Results
The Intelgic inspection system entered production and immediately delivered measurable improvements across detection accuracy, throughput, and quality consistency. The client eliminated the subjectivity that had long characterized their inspection process.
With each lens producing a timestamped, defect-mapped inspection record, the manufacturer gained traceability capabilities they had never previously possessed — enabling precise root cause analysis when defect trends emerged and providing documented quality evidence for regulatory and customer requirements.
Inspection throughput increased by more than six times compared to the manual process, enabling the production line to keep pace with increased order volume without adding inspection headcount. And because the AI operates with identical consistency across every shift and every operator, the quality floor — the minimum standard of inspection — is now defined by the AI, not by human variability.
"This is what a precision inspection system should look like — purpose-built fixtures, multi-condition imaging, and AI trained specifically on the defects that matter to optical performance."
Technology at a Glance
The complete system integrates Intelgic's machine vision AI with purpose-built hardware to create a cohesive, production-hardened inspection platform requiring minimal operator training and zero compromise on detection performance.
