Averroes Ai Automated Visual inspection software
PartnersCompany
Start Free Trial
Image
Image
Back
Defect Inspection

Macro Inspection Guide [Macro Defect Inspection Explained]

Logo
Averroes
Apr 20, 2026
Macro Inspection Guide [Macro Defect Inspection Explained]

Macro inspection sits at an awkward intersection.

It’s fast enough to run on every wafer, coarse enough that people underestimate it, and precise enough that when it’s misconfigured, the escapes are expensive. 

The spatial signatures it generates are some of the most diagnostically useful data in the fab. 

We’ll cover the full scope: defects, flow placement, implementation, and where AI is changing the capability ceiling.

Key Notes

  • 5 defect classes drive the majority of large-area yield loss and equipment damage events.
  • Recipe discipline and data integration determine whether macro inspection improves yield or generates noise.
  • AI reduces false positives from ~50% to single digits, but only with the right data infrastructure behind it.

What Is Macro Inspection?

Macro inspection is the automated (or semi-automated) imaging of entire wafers to detect large, visible defects – scratches, cracks, stains, chips, pattern excursions – that can kill multiple dies in a single event, cause wafer breakage, or damage expensive downstream equipment.

Macro Inspection vs. Micro Inspection vs. Metrology

These are three distinct layers of process control. 

They’re complementary – not interchangeable – and understanding the difference matters for designing an effective inspection strategy.

Attribute Macro Inspection Micro Inspection Metrology
Field of view Whole wafer Die / sub-die Site-specific
Magnification Low to intermediate High Varies
Primary goal Detect gross defects fast Characterize small defects Quantitative measurement
Defect scale ≥ 50 µm, large-area Single-digit micron to sub-micron N/A (measures dimensions)
Speed Fast – 100% wafer screening Slower, targeted Varies

The Practical Framing: 

Macro is a health screen; micro inspection and metrology are for root-cause analysis and process tuning at finer scales.

High-Resolution Micro Tools Can’t Substitute For Macro

Their field of view is too narrow and their throughput too low for rapid 100% wafer-surface screening. Metrology tools like CD-SEM or overlay measurement aren’t built for discrete defect detection at all. 

Each layer has a job – macro inspection’s job is fast, whole-wafer triage.

What Defects Does Macro Inspection Catch?

Macro defect inspection targets five broad defect classes. 

Together, they cover the failure modes most capable of causing catastrophic, multi-die yield loss or equipment damage in a single event.

1. Handling & Mechanical Damage

Scratches, cracks, edge chips, warpage, slip lines, delamination – front and back. 

Backside damage is the most dangerous variant: invisible from the front, capable of causing wafer breakage in CMP or litho. A single long scratch crossing multiple die rows can kill entire wafer regions in one handling event.

2. Surface Contamination & Stains

  • Particles, organic/inorganic films, residue, haze
  • Photoresist coverage issues and edge bead defects

Severity ranges from nuisance to catastrophic depending on particle size, location, and what tool comes next. 

A contaminated wafer entering a vacuum tool carries risk well beyond the wafer itself.

3. Lithography & Pattern Excursions

  • Reticle/stepper problems and flash-field defects
  • Gross pattern errors, ring-shaped or centered pattern anomalies
  • Incomplete liftoff, over/under-etch visible at macro scale

These are systematic by nature. One reticle defect stamps itself at the same location across every exposed die on the wafer – and potentially across the entire lot.

4. Edge & Bevel Defects

  • Edge carbonization and photoresist roll-off
  • Bevel contamination and edge-localized CMP or clean issues
  • Edge chips and cracks

The edge is the first surface to show process drift and the most common origin of mechanical failures. Many fabs underinvest here until a breakage spike forces the conversation.

5. Backside & Hidden Risks

  • Backside scratches, chips
  • Sub-surface slips appearing as long streaks or bands at macro scale

The defects most likely to cause catastrophic downstream loss – not because they’re difficult to detect with the right tool, but because they’re completely invisible to front-side-only inspection.

A Critical Point On Defect Character: 

Macro defects are predominantly systematic or excursion-driven, not random. 

A scratch pattern, a contamination ring, a reticle flash defect – these have spatial signatures that point directly to a specific tool, recipe, or material lot. 

That makes macro inspection not just a yield-protection step, but a rapid root-cause engine.

Where Macro Inspection Sits In The Fab Flow

Macro inspection is inserted at process transitions where the probability or downstream cost of a macro excursion is highest. 

The logic isn’t just yield, but also protecting the expensive equipment that comes next.

Key Insertion Points & The Rationale For Each:

Inline Vs. Stand-Alone Placement Is A Practical Trade-Off

  • Inline macro inspection (integrated into the track or CMP tool) catches problems faster and enables lot-hold before the next step. 
  • Stand-alone systems offer more flexibility (different magnifications, illumination modes, all-surface capability) but introduce handling and cycle-time considerations.

The Common Thread Across All Insertion Points: 

Macro inspection’s value is in early containment. Catching a defect at post-litho costs a fraction of catching it at test – and nothing compared to catching it in the field.

How Macro Inspection Works

Macro inspection systems image whole wafers using a combination of illumination modes at low-to-intermediate magnification, then apply automated defect classification (ADC) to identify, classify, and map defects.

Illumination Modes & What Each Reveals:

  • Brightfield: Reflected or transmitted light, best for surface films, stains, and pattern defects.
  • Darkfield: Scattered light at oblique angles, better for particles, scratches, and surface texture anomalies.
  • Oblique / multi-angle: Catches topographical features, edge defects, and low-contrast surface variations.

Most Capable Macro Tools Use Multiple Modes In Sequence

Layer-specific recipe tuning (not a single global recipe) is what separates effective macro inspection from systems that either miss defects or generate so much noise that operators stop trusting them.

Implementation Best Practices

The difference between a macro inspection program that improves yield and one that generates noise comes down to four areas:

Recipe Configuration

  • Tune brightfield/darkfield and magnification per layer and defect type – one global recipe will either miss defects or flood operators with false alarms.
  • Include intermediate magnifications to avoid blind spots between whole-wafer and micro-scale views.
  • Never ignore edge, bevel, and backside – many breakage events originate there.

Data Integration & Traceability

  • Tag every defect map with lot, wafer ID, layer, tool, and recipe. Without this, images can’t drive SPC, excursion detection, or root-cause analysis.
  • Feed macro data into centralized yield management, not siloed local GUIs.
  • Treat macro inspection data as process control data.

Recipe Governance

  • Assign clear ownership (yield/process engineering, with input from equipment and QA) and schedule periodic reviews after process changes or new product introductions.
  • Log recipe changes and correlate with false-positive and escape rates over time.
  • Don’t set-and-forget. As stacks, materials, and products evolve, static settings will either desensitize or generate alarm fatigue.

KPIs Worth Tracking:

Category Metric
Detection performance Capture rate, false-positive rate, escape rate
Operational Throughput (wafers/hr), time-to-detect excursion, review load per engineer
Business impact Wafer breakage rate, scrap/rework avoided, field returns from macro-level issues

AI & The Future of Macro Inspection

Rule-based ADC hits a ceiling. 

Recipe fragility, high false-positive rates, and inability to handle new defect types are the known failure modes. 

AI-based macro inspection addresses all three, but it introduces its own requirements.

What AI Adds To Macro Inspection:

  • Higher detection accuracy with lower false-positive rates
  • Faster recipe development through active learning and few-shot training
  • Detection of unknown or novel defect types that fall outside configured rule classes
  • Closed-loop feedback between inspection data and process adjustments

What AI Requires: 

  • Labeled training data
  • Rigorous validation
  • A retraining plan

AI model quality is tightly coupled to reference data quality and diversity. Deploying AI without this infrastructure doesn’t reduce false positives – it just changes where the errors come from.

What Does 99%+ Accuracy Look Like On Your Line?

Run AI on your existing equipment – no hardware changes needed.

 

Macro Inspection FAQs

What is the difference between macro inspection and AOI? 

Macro inspection and AOI (Automated Optical Inspection) are related but not the same thing. AOI is a broad category covering automated image-based inspection at various scales and steps. Macro inspection specifically refers to low-magnification, whole-wafer screening for large-area defects. In practice, many fabs run AOI tools in a macro inspection mode – but not all AOI is macro, and not all macro inspection platforms are traditional AOI systems.

What magnification is used in macro inspection? 

Macro inspection typically operates at low to intermediate magnification – from whole-wafer views up to approximately 10–20× for mid-scale defects. The exact range depends on the system and the defect classes being targeted. The key distinction from micro inspection is that macro tools prioritize field of view and throughput over resolution.

Can macro defect inspection be done on both sides of the wafer? 

Yes, modern macro defect inspection systems support all-surface inspection covering the frontside, backside, and bevel in a single pass. Front-side-only inspection leaves a significant blind spot, since backside scratches and chips are among the most common causes of catastrophic wafer breakage in CMP and lithography steps.

How does AI improve macro inspection accuracy on existing equipment? 

AI-based macro inspection improves accuracy by replacing static rule sets with models that learn from labeled defect examples – reducing false-positive rates from as high as 40–50% down to single digits in production deployments. Platforms like Averroes integrate directly with existing inspection equipment, no hardware changes required, and continuously improve detection performance through active learning as new defect types emerge.

Conclusion

Macro inspection doesn’t get enough credit for what it does. 

The fast, whole-wafer screen that prevents a cracked wafer from entering your scanner. The contamination ring that points straight back to a spin coater before the excursion hits a second lot. The backside chip that front-side-only systems never see.

The technology is moving – AI is replacing rule sets that were always going to break under process change – but the fundamentals don’t. Detect fast, contain early, read the signatures.

If your current macro inspection setup is generating more noise than signal, Averroes integrates with existing equipment, no hardware changes required. Book your free demo.

Related Blogs

Vision Systems for Final Product Quality Inspection and Defect Detection
Defect Inspection
Vision Systems for Final Product Quality Inspection and Defect Detection
Learn more
Glass Defect Detection & Inspection [2026 Guide]
Defect Inspection
Glass Defect Detection & Inspection [2026 Guide]
Learn more
KLA Defect Inspection: Comparing Bright-Field, Multi-Beam & E-Beam
Defect Inspection
KLA Defect Inspection: Comparing Bright-Field, Multi-Beam & E-Beam
Learn more
See all blogs
Background Decoration

Experience the Averroes AI Advantage

Elevate Your Visual Inspection Capabilities

Request a Demo Now

Background Decoration
Averroes Ai Automated Visual inspection software
demo@averroes.ai
415.361.9253
55 E 3rd Ave, San Mateo, CA 94401, US

Products

  • Defect Classification
  • Defect Review
  • Defect Segmentation
  • Defect Monitoring
  • Defect Detection
  • Advanced Process Control
  • Virtual Metrology
  • Labeling

Industries

  • Oil and Gas
  • Pharma
  • Electronics
  • Semiconductor
  • Photomask
  • Food and Beverage
  • Solar

Resources

  • Blog
  • Webinars
  • Whitepaper
  • Help center
  • Barcode Generator

Company

  • About
  • Our Mission
  • Our Vision

Partners

  • Become a partner

© 2026 Averroes. All rights reserved

    Terms and Conditions | Privacy Policy