Skip to main content

4.2 AOI Fundamentals

Automated opticalOptical inspectionInspection stands(AOI) systems function as the visual quality gate of the SMT line. Using high-resolution cameras and pattern-matching algorithms, AOI automatically inspects every board for component placement accuracy and solder joint integrity. The system's strategic placement (pre- or post-reflow) determines the cost-effectiveness of defect detection, as fixing a defect after reflow is significantly more expensive than correcting a placement error before the oven.

4.2.1 AOI Location and Strategic Placement

AOI can be deployed at multiple stages in the crossroadsSMT process, each offering distinct advantages for defect remediation.

Location

Purpose

Defects Caught

OpEx/CoQ Impact

Pre-Reflow

Verifies placement accuracy.

Missing components, Misalignment (X/Y), Rotation, Polarity errors.

High Value. Defects can be corrected by the PnP machine or an operator before the board enters the oven, preventing the permanent soldering of speed, consistency, and discernment in electronics manufacturing. By combining carefully chosen lighting, camera angles, and inspection logic, it provides a reliablebad checkassembly.

Post-Reflow

Verifies onfinal whetherjoint componentsquality.

Insufficient/Excessive areSolder, present,Bridging, aligned,Tombstoning, andSkewed visiblycomponents.

Verification. soldered to spec. Its strength lies in catching theCatches defects that showoccurred themselvesduring optically,the servingthermal asprocess. Leads to rework/scrap, which is expensive.

Strategy: Running a qualityPre-Reflow filterAOI is mandatory for complex or high-reliability boards. Catching a 0.4 mm BGA placement error before boardsreflow movesaves downstreamhours toof moreBGA expensiverework ortime.

destructive

4.2.2 tests.Defect TheCategories real challenge—and value—ofUpstream AOIRoot comes from balancing sensitivity with stability: catching every critical defect without drowning operators in false alarms.

4.2.1 What AOI is (and what it isn’t)Cause

AOI isuses design data (Gerber, CAD) as a fast,Golden consistentReference to check physical assemblies for deviations. Defects are categorized based on their impact:

AOI Defect Type

Upstream Failure Point

Consequence

Missing Component

Feeder failure (Chapter 2.3) or PnP miss-pick.

Open circuit, board failure.

Misalignment/Skew

PnP nozzle offset, board shift during placement, or paste slump (Chapter 1.5).

Poor wetting, tombstoning, potential shorts.

Polarity/Orientation

Kitting error or incorrect PnP programming (Chapter 2.2).

Component damage, circuit failure.

Bridging (Post-Reflow)

Excessive paste volume (SPI failure, Chapter 4.1) or severe misalignment.

Short circuit.

Insufficient Solder

Paste volume low (SPI failure) or poor wetting (Profile/Atmosphere issue, Chapter 3.3/3.4).

Weak joint, intermittent failure, open circuit.

4.2.3 2D vs. 3D AOI and Lighting Control

The core challenge for AOI systems is cameraFalse + lighting + software check that answers three questions:

  1. Is the right thing there?Calls (presence,False polarity, value markings)
  2. Is itPositives), where itglare shouldor be?shadow (X/Y/θincorrectly offset, lift/tilt)
  3. Does it look soldered? (bridges, insufficient/fillet shape, wetting clues)

It’s notflags a liegood detectorjoint foras everydefective. electricalThe fault.technology AOIchoice catchesdirectly visualaddresses risksthis early; ICT/FCT catch electrical or parametric faults later. Use both.issue.



4.2.2 Where to put AOI (pre- vs post-reflow)

  • Pre-reflow AOI (after placement): perfect for orientation/polarity, presence, wrong footprint, and big offsets before you bake mistakes in.
  • Post-reflow2D AOI: addsUses solderstandard joint judgement (bridges, opens, tombstones, fillet quality).
     Most lines run post-reflow as the main gate and use targeted pre-reflow checks for risky buildslighting (e.g., lotsdome, coaxial) to capture a flat image. This is fast and cost-effective but highly susceptible to reflections, often resulting in high false alarm rates.
  • 3D AOI: Uses structured light (e.g., laser triangulation or fringe patterns) to accurately measure height and volume of polarizedcomponent parts).bodies and solder fillets. This drastically reduces false calls by providing quantitative data beyond basic contrast matching.
  • Lighting Control: High-end AOI uses programmable multi-angle, multi-color LED arrays to eliminate shadows caused by tall components. Programming must be tuned to find a stable lighting recipe for each product.


OpEx
Impact:


Every false call requires an operator to stop, verify, and clear the alarm. A high False Alarm Rate (>5%) slows down throughput and increases OpEx faster than almost any other SMT issue. Investment in 3D AOI is typically justified by the reduction in these manual verification costs.

4.2.34 LightingProgramming &and angles—yourContinuous biggest quality knobsImprovement

ThinkAOI likeprogramming transforms the product's design data into a photographer:functional inspection routine.

  • Golden Board Reference: The Golden Board established during the wrongFirst lightArticle makes(Chapter good2.5) jointsserves lookas bad,the primary visual reference. The AOI program should be taught and vice-versa.optimized based on this known-good assembly.

  • Defect

    A)Library: LightingThe modesAOI system must maintain a comprehensive defect library, ensuring that unique flaws are classified correctly (picke.g., distinguishing a true short from a benign flux residue).

  • AI and Deep Learning: Modern systems leverage Artificial Intelligence (AI) to analyze thousands of images, teaching the fewestsystem thatto work)recognize acceptable variations (e.g., variations in solder fillet shape) while minimizing false positives, thus improving detection accuracy over time.
  • Gage R&R: Regular Gauge Repeatability and Reproducibility (GR&R) studies are required to verify the AOI system's ability to consistently provide the same result for the same defect. Low GR&R renders the AOI system unreliable.

Final Checklist: AOI Operational Mandates

ModeRequirement

WhatControl it highlightsPoint

UseQuality/Cost it forFocus

RingSystem / bright-fieldChoice

GeneralUse edges,3D silks,AOI texton lines with high component density and BGAs/QFNs.

Presence/offset,Minimizes OCV/OCR,OpEx polarityby marksdrastically reducing False Alarm Rate.

Low-angle / dark-fieldPlacement

Pre-Reflow AOITiny heightmust changes,be filletdeployed edgesfor high-risk assemblies.

Bridges,Catches liftedplacement/polarity leads,errors tombstonesbefore they become costly solder defects.

Coaxial (on-axis)Programming

Flat,Program reflectivemust surfaces

BGA masks, code reading on shiny parts

Structured/3D (fringe)

Height map of pads/parts

Lifted QFN corners, bent leads, solder mound height

B) Wavelength & color

  • White RGB covers most; blue can pop solder fillets; red tames mask glare; UV helps fluorescing conformal coats (post-coat lines).
  • Mask color/finish matters: OSP matte vs ENIG shiny will want different gains/angles. Save a lighting profile per product.

C) Camera geometry

  • Top camera (2D/3D) does the bulk of work.
  • Side cameras (oblique ~25–45°) see heel fillets on gull-wings, lifted QFN edges, and hidden bridges along tall parts.
  • If you have 3D AOI, use height to demote nuisance calls (e.g., silkscreen glare that 2D thinks is a bridge).



4.2.4 Rule-based vs ML libraries (and when to use which)

Approach

How it thinks

Strengths

Watch-outs

Rule-based (thresholds, geometry, color)

“Edges here, contrast there, pad X has area Y.”

Transparent, fast to tune; great for presence/offset/polarity and clear defects.

Brittle across finish/mask changes; many tiny rules to maintain.

ML-assisted (template/CNN classifiers)

Learns “this is OK vs NG” from examples.

Better at subtle fillet/joint quality, varied lighting, component cosmetics.

Needs curated training images; beware overfitting and hidden bias.

Practical blend: use rules for the must-haves (polarity, offset, bridges), and sprinkle ML on the subjective bits (fillet quality, cosmetic scratches). Keep ML outputs explainable (scores + example images).



4.2.5 Building a stable AOI program (simple recipe)

  1. Start from FA: teach on your Golden Board photos, not a random sample.
  2. Teach minimal features per part:
    • Presence box, pin-1/polarity locator, and pad windows for solder checks.
    • Avoid teaching glossy logos or changing lot codes as truth.
  4. Light sanity: lock exposure/gain per product. If you need more than two lighting modes on most parts, improve lighting before adding rules.
  5. Guard-band smartly: tighter on fine-pitch and polarity-critical parts; looser on cheap resistors that AOI will feed to process SPC anyway.



4.2.6 False calls vs escapes (how to balance)

  • False call = AOI flags a good feature → wastes touch time.
  • Escape = AOI misses a real defect → hurts yield/field.

Targets that keep lines calm (tune to your product):

  • False calls: ≤ 0.5–1.0 per board average, with a cap per panel.
  • Escapes: 0 on critical classes (polarity, bridges on high-risk nets); very low on others,be verified by audit sampling.

Levers to pull

  • Use 3D/side-view to reduce nuisance calls on fillets.
  • Create risk classes: Class A (polarity/bridges) = strict; Class B (cosmetics) = tolerant.
  • Route uncertain calls to a review queue with zoomed crops; don’t slowagainst the belt for a beauty contest.



4.2.7 Controlling drift (why yesterday’s good board fails today)

  • Mask/finish changes between lots change reflectivity → keep lighting profiles versioned by product and finish.
  • Lens/cover contamination raises false calls → clean optics on a schedule.
  • Board warp changes apparent fillet shape → fix supports upstream (Ch. 8.4) and let 3D judge height, not glare.
  • ECNs → treat AOI like code: rev the program when land patterns, silks, or part numbers change; attach the change note.



4.2.8 Metrics that matter (and ones to ignore)

Track per product, per side:

  • False calls/board (and by top 10 refdes)
  • Escapes found at ICT/FCT/rework (with AOI image back-link)
  • Top 5 AOI defect categories (bridges, tombstones, polarity, opens, cosmetic)
  • Review rate and auto-pass rate
  • Time to clear a board (don’t let AOI be the bottleneck)

Ignore “total defects counted” without context—it’s often just lighting noise.



4.2.9 Fast troubleshooting (AOI says NG—what now?)

  • Bridge calls spiking? Check stencil cleanliness and separation speed (7.5), then revisit lighting angle; don’t just widen thresholds.
  • Polarity NGs on LEDs/diodes? Verify silks & pin-1 marks are visible and consistent; switch to coaxial or add a simple OCR/OCV on the mark.
  • Fillet insufficient calls on gull-wings? Add side-view or 3D height; tune reflow TAL/peak slightly (9.2/9.4) before rewriting half the library.



4.2.10 Release checklist (stick this near the AOI)

  • Lighting profile saved (mode, angle, gains) for this product/finish
  • Program taught from Golden Board; pin-1/polarityuse checksmulti-angle on all polarized parts
  • Rule vs MLlighting splitto documented;control MLshadows.

    Ensures thresholdshigh showdetection scorescapability +(low example images

  • False callNegatives).

    Maintenance

    GR&R andstudies escapeperformed targetsregularly; set;AI Classmodel A defects strict, Class B tolerant

  • Optics clean, calibration check OK; side/3D cameras enabled where needed
  • Feedback loop active: escapes at ICT/FCT link back to AOI images/program rev



When AOI programs are built from Golden Board references, tunedretrained with propernew lightingdefect profiles,imagery.

Ensures the system remains reliable and balanced with risk-based thresholds, inspection shifts from noisy oversight to quiet reliability. The payoff is fewer escapes, less wasted touch-up effort, and a stable safeguard that keeps quality predictable without slowing production.


trustworthy.

Back to top