# DaoAI

> DaoAI (DaoAI Robotics Inc., Vancouver, BC, Canada) builds AI-native automated optical inspection (AOI) for PCBA. The AI auto-programs an inspection from one golden board in 30 seconds to 5 minutes, with no CAD file or component library required, and everything runs on-premise on the machine. The range covers offline 2D (P1, P1L, P2), inline 2D (P3, P3D), inline 3D (P5, P5D), available for delivery now, and through-hole (A3) machines.

## Key facts

- Every DaoAI AOI model programs itself from one golden board in 30 seconds to 5 minutes. No CAD, Gerber, centroid file or component library is required.
- Lineup by type: offline 2D (P1, P1L, P2), inline 2D (P3, P3D), inline 3D (P5, P5D) and through-hole (A3).
- The inline 3D AOI machines P5 and P5D are available for delivery now.
- The A3 inspects through-hole (THT) assemblies both pre- and post-wave in one machine.
- All AI runs on-premise on the machine; inspection data does not leave the factory.
- Sales and engineering support in North America comes from the DaoAI team in Vancouver, BC, plus authorized partners.
- DaoAI is recruiting distributors and sales reps in the US, Canada and Mexico: https://daoai.com/distributors
- Meet DaoAI at SMTA International 2026, Booth 2727, October 27–29, 2026, Rosemont, IL: https://daoai.com/events/smta-international-2026

## Machines

- [AOI Machines for PCB Assembly: Compare All Models](https://daoai.com/machines): Compare DaoAI offline 2D, inline 2D, inline 3D and through-hole AOI machines. Every model programs itself from one golden board in 30 s to 5 min, with no CAD.

### P1: Offline AOI Machine

URL: https://daoai.com/machines/p1

Manual-load offline AOI for production inspection and lab use. No conveyor required, with the same core AI software as the rest of the range.

P1 offline AOI machine with a 5 MP camera at 15 µm per pixel for boards up to 470 × 330 mm. Manual load and AI auto-programming from one golden board, no CAD.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 5 MP |
| Resolution | 15 µm per pixel |
| Lens | Dual telecentric |
| Speed | 0.238 sec per FOV |
| Lighting | Multi-spectral high-speed RGB, 6 channel |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 470 × 330 mm |
| Board thickness | 0.5 – 5 mm |
| Max. PCB clearance | Top 25 mm; Bottom 110 mm |
| Board handling | Offline, manual load. Auto-clamp |
| Line integration | None. Offline machine |

#### Inspection capability

| Spec | Value |
| --- | --- |
| Two-dimensional items | 16 named inspection items, from missing component to insufficient solder |
| Through hole | Through-hole joint and bridge, judged per pad |
| Marking | AI OCR for unknown markings, AI OCV for verification against an expected string |
| Panel | Rows and columns found automatically, bad board marks honoured |
| Whole board | Not available on P1 |

#### Additional functions

| Spec | Value |
| --- | --- |
| Programming | AI auto-programming from one golden board. CAD optional |
| Offline programming | Program away from the line and move the program to the machine |
| Model training | On-machine, offline. Your images never leave the plant |
| Analysis | Gauge R&R, Cpk, sub-panel first pass yield, all exportable to Excel |
| MES export | Defect codes mapped to your vocabulary, with cropped defect images |
| Repair station | Separate screen for the repair operator |
| Accounts | Admin and operator roles. Hand-tuned parameters protected from regeneration |
| Remote support | Access opened by you, closed by you |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

### P1L: Large-Format Offline AOI

URL: https://daoai.com/machines/p1l

A larger inspection area for boards that exceed the P1’s capacity. Manual loading, with the same core AI software.

P1L large-format offline AOI for boards up to 680 × 510 mm. 5 MP camera at 15 µm per pixel, manual loading, and AI auto-programming from one golden board.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 5 MP |
| Resolution | 15 µm per pixel |
| Lens | Dual telecentric |
| Speed | 0.238 sec per FOV |
| Lighting | Multi-spectral high-speed RGB, 6 channel |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 680 × 510 mm |
| Board thickness | 0.5 – 5 mm |
| Max. PCB clearance | Top 25 mm; Bottom 110 mm |
| Board handling | Offline, manual load. Auto-clamp |
| Line integration | None. Offline machine |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

### P2: Offline AOI · 12 MP Flagship

URL: https://daoai.com/machines/p2

A 12-megapixel camera provides a resolution of 10 µm per pixel, with the same board capacity and footprint as the P1.

P2 flagship offline AOI with a 12 MP camera at 10 µm per pixel for boards up to 470 × 330 mm. Same footprint as the P1, same AI auto-programming software.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 12 MP |
| Resolution | 10 µm per pixel |
| Lens | Dual telecentric |
| Speed | 0.238 sec per FOV |
| Lighting | Multi-spectral high-speed RGB, 6 channel |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 470 × 330 mm |
| Board thickness | 0.5 – 5 mm |
| Max. PCB clearance | Top 25 mm; Bottom 110 mm |
| Board handling | Offline, manual load. Auto-clamp |
| Line integration | None. Offline machine |

#### Inspection capability

| Spec | Value |
| --- | --- |
| Two-dimensional items | 16 named inspection items, from missing component to insufficient solder |
| Through hole | Through-hole joint and bridge, judged per pad |
| Marking | AI OCR for unknown markings, AI OCV for verification against an expected string |
| Panel | Rows and columns found automatically, bad board marks honoured |
| Whole board | Solder balls and foreign objects between components |

#### Additional functions

| Spec | Value |
| --- | --- |
| Programming | AI auto-programming from one golden board. CAD optional |
| Offline programming | Program away from the line and move the program to the machine |
| Model training | On-machine, offline. Your images never leave the plant |
| Analysis | Gauge R&R, Cpk, sub-panel first pass yield, all exportable to Excel |
| MES export | Defect codes mapped to your vocabulary, with cropped defect images |
| Repair station | Separate screen for the repair operator |
| Accounts | Admin and operator roles. Hand-tuned parameters protected from regeneration |
| Remote support | Access opened by you, closed by you |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

### P3: Inline AOI Machine · Single Track

URL: https://daoai.com/machines/p3

2D inline AOI for boards up to 500 × 600 mm. AI auto-programming starts from one golden board.

P3 single-track inline AOI machine with a 26 MP camera at 10 µm per pixel for boards up to 500 × 600 mm. AI auto-programming starts from one golden board.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 26 MP, global shutter |
| Resolution | 10 µm per pixel, 51 × 51 mm field of view |
| Lens | Telecentric |
| Lighting | RGB coaxial ring, multi-angle LED |
| Depth of field | 8 – 10 mm |
| Inference | Under 1 ms per component |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 500 × 600 mm |
| Board thickness | 0.3 – 6 mm |
| Clearance | Top 30 mm, bottom 20 mm. Custom on request |
| Smallest component | 0201, 0.3 mm pitch ICs. 01005 optional |
| Board handling | Inline, single track. Conveyor height 900 ± 25 mm |
| Double sided | Top and bottom side switching, automatic |

#### Inspection capability

| Spec | Value |
| --- | --- |
| Two-dimensional items | 16 named inspection items, from missing component to insufficient solder |
| Through hole | Through-hole joint and bridge, judged per pad |
| Marking | AI OCR for unknown markings, AI OCV for verification against an expected string |
| Panel | Rows and columns found automatically, bad board marks honoured |
| Whole board | Solder balls and foreign objects between components |

#### Additional functions

| Spec | Value |
| --- | --- |
| Programming | AI auto-programming from one golden board. CAD optional |
| Offline programming | Program away from the line and move the program to the machine |
| Model training | On-machine, offline. Your images never leave the plant |
| Analysis | Gauge R&R, Cpk, sub-panel first pass yield, all exportable to Excel |
| MES export | Defect codes mapped to your vocabulary, with cropped defect images |
| Repair station | Separate screen for the repair operator |
| Accounts | Admin and operator roles. Hand-tuned parameters protected from regeneration |
| Remote support | Access opened by you, closed by you |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

Capability: DIP / through-hole inspection (https://daoai.com/features/dip).

### P3D: Dual-Track Inline AOI Machine

URL: https://daoai.com/machines/p3d

Dual-track 2D inline AOI for high-throughput production, with the same optics and core AI software as the P3.

P3D dual-track inline AOI for high-throughput SMT lines: 26 MP camera at 10 µm per pixel, boards up to 500 × 325 mm, same optics and AI software as the P3.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 26 MP, global shutter |
| Resolution | 10 µm per pixel, 51 × 51 mm field of view |
| Lens | Telecentric |
| Lighting | RGB coaxial ring, multi-angle LED |
| Depth of field | 8 – 10 mm |
| Inference | Under 1 ms per component |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 500 × 325 mm, per track |
| Board thickness | 0.3 – 6 mm |
| Clearance | Top 30 mm, bottom 20 mm. Custom on request |
| Smallest component | 0201, 0.3 mm pitch ICs. 01005 optional |
| Board handling | Inline, dual track. Conveyor height 900 ± 25 mm |
| Double sided | Top and bottom side switching, automatic |

#### Inspection capability

| Spec | Value |
| --- | --- |
| Two-dimensional items | 16 named inspection items, from missing component to insufficient solder |
| Through hole | Through-hole joint and bridge, judged per pad |
| Marking | AI OCR for unknown markings, AI OCV for verification against an expected string |
| Panel | Rows and columns found automatically, bad board marks honoured |
| Whole board | Solder balls and foreign objects between components |

#### Additional functions

| Spec | Value |
| --- | --- |
| Programming | AI auto-programming from one golden board. CAD optional |
| Offline programming | Program away from the line and move the program to the machine |
| Model training | On-machine, offline. Your images never leave the plant |
| Analysis | Gauge R&R, Cpk, sub-panel first pass yield, all exportable to Excel |
| MES export | Defect codes mapped to your vocabulary, with cropped defect images |
| Repair station | Separate screen for the repair operator |
| Accounts | Admin and operator roles. Hand-tuned parameters protected from regeneration |
| Remote support | Access opened by you, closed by you |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

Capability: DIP / through-hole inspection (https://daoai.com/features/dip).

### P5: Inline 3D AOI Machine

URL: https://daoai.com/machines/p5

The 3D machine on a single-track conveyor. The same AI inspection, with component height measured rather than inferred.

P5 inline 3D AOI machine with 4-direction projection and height accuracy of 6 µm or better with the calibration tool, for boards up to 500 × 600 mm.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 21 MP high-speed industrial camera |
| Resolution | 10 µm optical, 50 × 40 mm field of view |
| 3D sensor | Four Scheimpflug structured-light projection units, one per direction |
| Height resolution | ≤ 6 µm height accuracy with the calibration tool. 25 mm measurement range |
| Speed | Over 10 cm² per second |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 500 × 600 mm |
| Board thickness | 0.6 – 5 mm |
| Component height | Top 50 mm, bottom 40 mm |
| Smallest component | 0603 metric, 0201 inch |
| Board handling | Inline, single track. Conveyor height 900 ± 20 mm |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

Capability: DIP / through-hole inspection (https://daoai.com/features/dip).

### P5D: Dual-Track Inline 3D AOI

URL: https://daoai.com/machines/p5d

Dual-track 3D AOI for high-throughput production, with the same optics and software as the P5.

P5D dual-track inline 3D AOI with 4-direction projection and height accuracy of 6 µm or better with the calibration tool, for boards up to 500 × 300 mm.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 21 MP high-speed industrial camera |
| Resolution | 10 µm optical, 50 × 40 mm field of view |
| 3D sensor | Four Scheimpflug structured-light projection units, one per direction |
| Height resolution | ≤ 6 µm height accuracy with the calibration tool. 25 mm measurement range |
| Speed | Over 10 cm² per second |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 500 × 300 mm, per track |
| Board thickness | 0.6 – 5 mm |
| Component height | Top 50 mm, bottom 40 mm |
| Smallest component | 0603 metric, 0201 inch |
| Board handling | Inline, dual track. Conveyor height 900 ± 20 mm |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

Capability: DIP / through-hole inspection (https://daoai.com/features/dip).

### A3: Through-Hole (THT) AOI

URL: https://daoai.com/machines/a3

AI inspection for through-hole assemblies before and after wave soldering. Alongside standard THT solder joint and bridge checks, dedicated AI algorithms detect missing, incorrect and reversed components.

A3 through-hole (THT) AOI for pre-wave and post-wave inspection: 26 MP camera, boards up to 420 × 420 mm, AI checks for missing, wrong and reversed parts.

#### Optics

| Spec | Value |
| --- | --- |
| Camera | 26 MP colour industrial camera |
| Lens | 18 mm industrial imaging lens |
| Lighting | High-brightness LED open-aperture flat light plus low-angle light |

#### Conveyor and board

| Spec | Value |
| --- | --- |
| Max board size | 50 × 50 mm – 420 × 420 mm |
| Board handling | Bench-top straddle over the existing line, no line modification. Height 1435 – 1635 mm and width 1000 – 1200 mm adjustable |
| Machine depth | 841 mm |
| Weight | 100 kg |
| Power | 220 V, 10 A, 50 – 60 Hz |

Programming input: One golden board — no CAD, Gerber, centroid file or component library required.

Hardware specifications are reference values and may vary by configuration.

## Features

### AI Auto-Programming

URL: https://daoai.com/features/auto-programming

AI AOI auto-programming from one golden board: every component detected, every window and threshold set in 30 seconds to 5 minutes. No CAD or library needed.

- 30 s – 5 min — To generate an inspection program from scratch — Achieve a 1 % false call rate with tuning.
- Detection accuracy, in one click — Straight out of auto-programming, before any tuning.
- CAD optional — One golden board is the whole input — A CAD or BOM file fills in designators and part numbers. Nothing else depends on it.
- Days — None — No waiting for data — Nothing starts until CAD or a component list arrives, and is correct.
- Less programming time per board — Programming a new board takes an engineer 3 to 5 hours, and several days on a board with thousands of components.
- Prerequisite — None — No component library required — Every new package added by hand, and kept correct from then on.
- Hours — Changeovers start ready — A new board means programming again, with thresholds tuned one at a time.
- Components — Missing, wrong, shifted, rotated or damaged components, and tombstoning.
- Component polarity — Reversed diodes, LEDs, capacitors, tantalums and ICs.
- Component markings — Wrong component codes, missing characters, smudged or shifted print.
- Component orientation — Components placed the right way up but turned 90 or 180 degrees.
- Solder quality — Insufficient or excess solder, cold joints, and bridges between neighbouring pads.
- Pins, pads and solder — Lifted, bent, damaged or missing leads, one verdict per pin.
- Through-hole joints — Before the wave: missing, reversed or badly inserted components. After the wave: insufficient or excess solder, bad or bridged joints.
- Red glue — Adhesive that has spread onto a pad before wave soldering.
- Board + CAD/ BOM
- One Golden Board Only
- Programming quality — Full. The file does not change what is detected or how it is judged — Identical. Components, leads and pads are all detected from the board
- When you can start — As soon as a golden board is available. Import CAD/BOM data whenever it is available. — As soon as a golden board is available.
- Designators and part numbers — Your real designators and part numbers, matched in automatically — Auto-generated identifiers. Positions are correct but unnamed
- Grouping across identical components — Grouped by part number as well, so one edit covers the exact component — Grouped by detected component type
- MES and traceability — Exports in the vocabulary your MES already uses, no mapping — Exports with generated identifiers, mapping needed downstream
- Operator and repair station — Defect named by designator, which is what the repair sheet says — Defect located by position on the board image
- Accepted formats are the component information files SMT equipment already exports, in CSV, TSV or plain text. If one arrives late, or never, inspection quality is unaffected.
- No pre-built component library is required for auto-programming. Component templates can be saved, reused and exported to other machines.
- STEP 01 — Capture and choose — Capture the golden board, then tick the checks you want.
- STEP 02 — Auto-program — The AI generates the inspection program in 30 seconds to 5 minutes. For panelized boards, it also generates the panel layout in the same run.
- STEP 03 — Fine-tune — Change anything you are not happy with.
- STEP 04 — Run it — Start inspecting production boards. Operators can review results and correct false calls directly at the machine.
- Adjust bounding boxes — Set detection bounding box sizes for each package type before auto-programming, instead of resizing them component by component afterward.
- Choose inspection regions — For each package type, choose whether to skip a region, generate its geometry only, or include it in inspection. This includes pad, pin and bridge regions.
- Lock selected parameters — Lock a parameter to your chosen value so it stays unchanged when you run auto-programming again.
- Save as presets — Save the whole auto-programming setup under a name, apply it to the next product in one step, or export it to another machine.
- Rule-based teach — TEMPLATE MATCHING · GOLDEN SAMPLE
- Generic AI AOI — PRE-TRAINED MODEL · CLOUD PROCESSED
- DaoAI — AUTO-PROGRAMMING
- What it needs before you start — CAD or a component list, plus a package library somebody maintains — A pre-trained model that was never trained on your boards — One golden board — CAD is accepted and optional
- How a component is judged — Pixel by pixel against one golden image — A general model trained on other people's boards, never fitted to yours — Components, not pixels — An AI model trained on over a million real production images. The model is adapted to your board to help distinguish defects from acceptable visual variation.
- A new component type — Someone adds it to the library by hand first — Works only if it resembles the training data. Otherwise, back to the vendor — Nothing to add first — Detected and programmed straight from the board
- Bringing the false call rate down — Loosen the threshold, which also lets real defects through — Raised as a request. The improvement arrives when the vendor retrains and ships it — Mark it OK once — The model retrains on it. See Feedback Learning
- Who is able to change it — Whoever holds the threshold knowledge — Their AI engineer. Cloud processing also carries your board images outside the plant — Any operator — Can see what to change and how. Account permissions decide who is allowed to
- Does it need CAD or Gerber files? — No. One golden board is enough. CAD is accepted, and it adds designators, part numbers and grouping. It does not change programming quality. — See the CAD comparison above.
- What happens on a package the system has never seen? — It is detected and modelled from the board in front of it. There is no library entry to create first, which is the point of the whole approach.
- If I tune a parameter by hand, will retraining overwrite it? — No. Generation compares when the parameter was last edited by hand against when its region was last generated, and it leaves the newer human edit alone.
- How accurate is it before we tune anything? — 98% detection accuracy on the program the machine writes itself, in one click. Reporting false calls from the review screen takes it further from there.

### Feedback Learning

URL: https://daoai.com/features/feedback-learning

Reduce AOI false calls with feedback learning: operators report a false call or a missed defect, and the AI learns the right parameters from those reports.

- One click — From a wrong call to a training example — Reported at the review screen, during production. Nothing leaves the machine.
- Fewer false calls — Improves with operator feedback — The model learns from reviewed calls and applies that feedback in later inspections.
- Continuous learning — Improve the model with each round of feedback. — Update after reviewing a false call, or collect feedback during production and update later.
- STEP 01 · OPERATOR — Mark it OK or NG — Mark the component as OK or NG. Your feedback can apply to other components in the same group.
- STEP 02 · OPERATOR — Choose what the AI learns from — Turn training on to include this feedback in model updates. Leave it off to record the feedback without using it for training.
- STEP 03 · THE AI — Update the AI — Run AI Update to learn from the selected feedback and adjust inspection parameters automatically.
- STEP 04 · MACHINE — Re-inspect — Inspect the board again and review the updated results.
- Your settings stay — An update works around anything you set yourself.
- Manual settings take priority — Where your setting and the update disagree, yours is kept.
- The line comes first — An update waits for the line, not the other way around.
- Undo an update — Not what you wanted? Put it back the way it was.
- What actually changes — The accept band widens for everything judged by that parameter — Nothing on your machine. The request goes to the vendor and the deployed model stays as it is — The model learns — It gains an example of what that component looks like when correct
- Effect on real defects — Defects inside the widened band are accepted too — Unknown until the next model arrives, and it was not trained on your defect — Still caught — A defect near the same score is still outside what is now normal
- Where the decision lives — A number somebody changed, and the reason is not stored with it — With the vendor, in a dataset you do not hold — In your dataset — A labelled example in the product’s dataset, with the image attached
- Scope of the change — Everything that parameter governs, whether you meant it or not — Whatever their retraining changed, across every plant running that model — The group you choose — One position, a class, a part number, or a package
- Who can do it — Whoever has permission to change inspection parameters — An AI engineer, usually theirs — Any operator — Calls can be reported without permission to change the program
- Long term — The threshold stays loose and the reason is not recorded — You wait on a release cycle, and your images may have to leave the plant to get there — Fixed for good — The model has one more example and does not call that component again
- How many reports before it stops calling that component? — It depends on the item and how different the component looks. The machine tells you when a product has collected enough to be worth an update.
- Does reporting a false call change the program? — No. A report adds an example. The program changes when someone with that permission runs an update, and that can be undone.
- Does the line stop while it retrains? — No. An update waits for the line rather than the other way around.
- Will an update overwrite our engineer’s thresholds? — No. Anything you set by hand is what the machine keeps.
- Tune thresholds — BY HAND
- Generic AI AOI — VENDOR RETRAINS AND SHIPS IT
- DaoAI — FEEDBACK AND AI UPDATE

### Model Training

URL: https://daoai.com/features/model-training

Train custom AOI defect models offline, on the machine, with your own images. Built for defects specific to your product that no AOI ships an inspection for.

- Add to the dataset — One click on any inspected region, or a whole product at once, puts the images into the dataset. Feedback from the review screen lands there too.
- Label OK or NG — Label each image as OK or NG. The software tracks unlabelled images so you can pause and resume later.
- Train — Name your model, start training and monitor its performance. Stop at any time and keep the best result so far.
- Check the result — Evaluate the model on images not used for training. Review correct predictions, false calls and missed defects in a results table.
- Apply the model — Apply the model to selected inspection regions and set it as the default for the group. It runs alongside your existing inspection checks.
- No internet, no cloud, no account — Train locally on the machine, without an internet connection, cloud service or online account.
- Your images stay local — Collect, label and train on your images locally. No image upload is required.
- The model is your property — Your team trains the model on your production images. The model belongs to you and is not shared with other customers.
- Train separately from production — Collect and label images, train on a separate machine, then transfer the finished model to your production machine.
- Add to your existing inspections — Your custom model adds a check to selected inspection regions. Existing inspection checks continue to run.
- Update or replace your model — Disable the model for a group, retrain it with more examples or replace it with a newer version.

### Full Control

URL: https://daoai.com/features/full-control

AOI program control and threshold tuning: 21 settings across six areas stay yours to change, from one threshold to what gets generated, and survive every run.

- Set It First — Configure before auto-programming — Choose which inspection regions and checks to generate before running auto-programming.
- Change Once — Applies to every component of that kind — Apply a change to one position, a part-number group or an entire package type. You choose the scope.
- Your Settings Stay — Manual settings are preserved — Your manual settings are retained when you rerun auto-programming.
Four levels of control.
Override the AI at exactly the scope you need.
Adjust any threshold
Adjust AI-generated thresholds and review how the changes affect inspection results. Define the range available to AI updates, or disable AI updates for that check.
Set the rules for a whole package type
Set inspection rules for each package type instead of editing components individually. Save them as presets to reuse on other products.
Add a component or inspection region
Draw an inspection box around a component or feature that was not detected. The system learns its parameters from the golden board and adds it to the inspection program.
Train your own model
Train a custom model on your own images for niche defects specific to your products, then apply it to selected inspection regions.
Model Training →
- Generation presets — For each package type, choose how pad, pin, solder and bridge regions are handled: generate and inspect, generate only, or skip. — cover — Generation presets: the checks auto-programming makes for each package type, set per package.
- Parameter presets — Save default inspection parameters as named presets. Reuse them across products or export them to other machines for consistent programming. — contain — Parameter presets: the default parameters for every check, saved as a named preset.
- What it generates — - Inspection box size — Preset box sizes for each package type. — - Pad, pin and bridge regions — Generate and inspect, generate only, or skip. — - Sensitivity — Set as a percentage, per check. — - Alignment marks — Placed for you, or click the centres yourself.
- How it judges — - Any threshold — Adjust while reviewing inspection measurements. — - Locked values — Keep selected parameter values fixed. — - Which inspections run — Enable or disable checks by package type. — - Judgement criteria — Edit solder joint acceptance limits.
- Where it looks — - Your own regions — Draw a box anywhere on the board. — - What runs inside it — Choose the inspection for a region you drew. — - Panel layout — Draw one region, every sub-panel is found. — - Region geometry — Move and resize anything that was generated.
- What it learned — - Report a false call — From the review screen, during production. — - What goes into training — Remove any report from the dataset before training. — - Your own model — Collect, label, train, check, apply. — - Turn a model off — For one group, without affecting the rest.
- What things are called — - Designators and part numbers — From a CAD file, or typed in. — - Grouping level — One position, a part number, a package, a class. — - Ambiguous characters — Set equivalent characters, such as 0 and O.
- Who may do it — - Permission per level — Program, tune and train—or inspect only. — - Undo — Revert supported changes.
Your manual settings take priority.
Auto-programming and model updates preserve your manual settings. Locked parameters remain fixed at the values you choose.
- Do I have to fix regions one at a time? — No. Set inspection rules for a package type to apply them across the group. Save the rules as presets to reuse on other products.
- If I change something, will the AI change it back? — Auto-programming and model updates preserve your manual settings. Locked parameters remain fixed at the values you choose.
- Can I disable solder or bridge inspection for a specific package type? — Yes. Choose whether the corresponding regions are generated and inspected, generated without inspection, or skipped.
- Can we train a model for product-specific defects? — Yes. Train a custom model on your own images. See Model Training for the collection, labelling, training and evaluation workflow.

### Inspection Coverage

URL: https://daoai.com/features/inspection-coverage

AOI inspection coverage across 16 inspection capabilities, each run on an automatically generated program. Select the checks that apply to each package type.

- Component inspection — 4 CHECKS — - Component Presence & Identification — Missing or incorrect components. — - Placement, Orientation & Polarity — X/Y offset, skew or rotation, incorrect polarity, tombstoning, billboarding, and upside-down placement. — - Component Condition — Package or body damage, damaged chip terminations, burn marks, and discoloration. — - Component Height & Coplanarity — Measures component height and detects lifted or tilted component bodies and coplanarity defects using 3D AOI.
- Solder joint & lead inspection — 5 CHECKS — - Solder Joint Quality — Insufficient solder, excess solder, and cold solder joints. — - Open Solder Joints — Open or unsoldered connections, available in both 2D and 3D configurations. — Explore open-solder detection → — - Lead Condition — Lifted, bent, damaged, missing, or unsoldered component leads. — - Solder Bridging — Unintended solder connections between adjacent pads or leads. — - Adhesive Contamination on Pads — Unwanted SMT adhesive (“red glue”) on solder pads.
- Marking & code inspection — 2 CHECKS — - OCR / OCV — Reads and verifies alphanumeric component markings, symbols, and laser-marked text. — - Barcode & Serial Number Reading — Reads 1D barcodes, QR codes, other 2D codes such as Data Matrix, and serial numbers for unit-level traceability.
- Through-hole inspection — 2 CHECKS — - Through-Hole Lead & Solder Joint Inspection — Pre-wave: lead/pin presence and insertion. Post-wave: missing, insufficient, or excess solder. — - Through-Hole Solder Bridging — Solder bridges between adjacent through-hole pads.
- Whole-board inspection — 3 CHECKS — - Solder Balls & Particles — Solder balls or particles anywhere on the assembly, not only adjacent to a solder joint. — - Foreign Object Debris (FOD) — Debris, wire clippings, dropped components, and other foreign material. — - Surface Contamination — Unacceptable flux residue, stains, or other surface contamination. Coming soon.

### Panel Inspection

URL: https://daoai.com/features/panel

Panel and array AOI: program one board and inspect the whole panel. Rows, columns and spacing come from the image, and every sub-panel is reported on its own.

- One click — Generate the whole panel — SECONDS, ONCE PER PANEL — Draw a box around one sub-panel and run auto-detection to find the remaining positions. — - No rows, columns or spacing to enter. — - Matching components across sub-panels provide training examples for the same model. — - Apply shared edits across matching positions without repeating the work.
- One click — Align every position — AUTOMATIC ALIGNMENT — Align inspection regions to each sub-panel to compensate for position and spacing variations across the panel. — - Each sub-panel is aligned against its own features, not the panel as a whole. — - Help reduce false calls caused by misalignment. — - Re-align after a fixture change without touching the inspection program.
- Edit one position independently — Edit one position independently while the others retain the shared program. Your manual settings are preserved during regeneration.
- Handle irregular layouts — Configure irregular spacing or mixed orientations manually, then save the layout for that product.
- Reuse a program on a new layout — Reuse an existing sub-panel program on a different panel layout without programming it again.
- Draw one sub-panel. The AI finds the rest. — Draw a box around one sub-panel. The system identifies the repeated board positions and detects the panel layout automatically.
- Traceable after depaneling — Link each sub-panel’s barcode or serial number to its inspection results, so individual boards remain traceable after depaneling.
- Automatic orientation detection — The AI detects each sub-panel’s orientation from the image, without manual angle entry.
- One program, every position — Create the inspection program on one sub-panel, then apply it across the panel. Automatically align the inspection regions at each position.

### Whole-Board Inspection

URL: https://daoai.com/features/whole-board

Solder ball and foreign object detection AOI: a whole-board pass inspects the bare laminate between components, where component inspection windows never look.

- Solder Balls & Particles — Solder balls or particles anywhere on the assembly, not only adjacent to a solder joint.
- Foreign Object Debris (FOD) — Debris, wire clippings, dropped components, and other foreign material.

### Open Solder Detection

URL: https://daoai.com/features/open-solder

Open solder joint detection with a pretrained 2D AI model. Detect visible open-solder defects on PCBA, then fine-tune locally with samples from your production.

Open Solder Detection
Detect open solder joints.
With 2D AI.
DaoAI identifies visible open-solder defects using a pretrained AI model and 2D images. Use the supplied model directly, then fine-tune it with a small set of production samples when needed.
See detection examples
Open-solder detection in a 2D inspection image.
More than a color threshold.
Color alone does not tell the whole story of a solder joint. Reflections and surface variation can make fixed color rules difficult to apply consistently.
- Traditional color-based inspection — Conventional color-based solutions judge solder joints against preset color ranges. But a change in color is not necessarily a defect, and a color match does not guarantee a good joint. This can lead to false calls, missed open-solder defects, and repeated threshold tuning.
- Trained on the defect. — DaoAI uses a model trained on representative open-solder defects to recognize visual patterns in 2D inspection images. The decision is not based on color limits alone.
Two joints. Only one model sees the difference.
Both are chip terminations in a 2D inspection image: one lead is soldered to its pad, the other is not. A color range set wide enough to pass the joint on the left passes the open joint on the right as well. That is where color-based inspection stops.
- Acceptable joint — - Other model — Pass — - DaoAI model — Pass
- Open solder joint — - Other model — Pass — - DaoAI model — Open solder
DaoAI model
An open solder joint lacks a soldered connection between a component lead or termination and its pad. DaoAI identifies visible signs of these defects in 2D inspection images.
- Blog — Why color thresholds can miss defects
- Blog — Read the Open Solder guide
Use the supplied model. Fine-tune when needed.
- Ready to use. — Start with the supplied model. You do not need to train a new model from scratch before your first inspection.
- Adaptable to your boards. — When your boards or process conditions differ, collect a small set of representative samples from your line and fine-tune the model locally for your application.
- Feature — Explore model training
- Feature — See feedback learning
Model-training workflow.
See open-solder detection in action.
Talk with a DaoAI engineer about your solder-joint samples and inspection requirements.
Book a demo
What would you like to inspect?
Tell us about your board, component types, and solder-joint defects.
Next in Detection
Inspection Coverage →
Open solder is one defect type. Inspection Coverage lists every defect the machine looks for on a board, and where each one is checked.

### OCR & OCV

URL: https://daoai.com/features/ocr

AOI OCR and OCV for component markings: OCR reads print it was not told about, like a serial number; OCV confirms the expected marking is there and correct.

- OCR — Recognition — Read it — What does this say? — var(--border) — - Use it when — Read serial numbers, date codes, lot codes and component markings without entering the expected text first. — - What you get back — Recognized text saved with the inspection results, including serial numbers for unit-level traceability. — - Handling similar characters — Configure character equivalence groups to help reduce false calls caused by similar-looking characters.
- OCV — Verification — Check it — Does this say what it should? — var(--accent) — - Use it when — Verify known text, such as component codes, designators and fixed board markings. — - What you get back — Verification results showing whether the expected text is present, correct and legible. — - Why use verification — OCV checks against the expected text rather than identifying each character without a reference.
- Zero and the letter O can look alike.
- Similar strokes can make these characters hard to distinguish.
- Print quality can obscure the difference between 5 and S.
- Ink spread can make the shapes of 8 and B look similar.
- Component markings — Read component codes, date codes and manufacturer markings directly from the component body.
- Board silkscreen — Read designators and other text printed on the board.
- Barcodes and 2D codes — Read 1D barcodes, QR codes and other 2D codes such as Data Matrix on boards or panels during inspection.
- Serial numbers — Save each unit’s serial number with its inspection results for traceability.

### Through-Hole Inspection

URL: https://daoai.com/features/dip

Through-hole AOI inspection before and after wave soldering: check insertion, polarity and position pre-wave, then every solder joint post-wave, in one program.

- Collect images from inspected boards — Add individual solder joint images or all matching solder joint images from a product to your training dataset.
- Label OK and NG to your standard — Label solder joint images as OK or NG to teach the model your acceptance criteria.
- Apply the model where needed — Apply the trained model to selected pads or set it as the default for a component group. It runs alongside the existing joint and bridge checks.

### Production Integration

URL: https://daoai.com/features/in-your-plant

AOI production integration with MES export, Cpk quality analysis, a dedicated repair station screen and line handling, all built into the inspection software.

- Line handling — 3 CAPABILITIES — - Capture Path Optimization — The camera stops only where the board requires it. Fewer captures per board with every check enabled. — - Top/Bottom Side Program Selection — Program each side of a double-sided assembly separately, then run both in one inspection. The AI recognises which side is in front of the camera and applies the matching program. — - Pass/Fail Board Routing — Failed boards are sorted according to the configured strategy, without operator intervention.
- Quality analysis — 3 CAPABILITIES — - Gage R&R (GR&R) — Repeatability study on placement accuracy (X offset, Y offset, rotation), run in the software on your boards and reported as precision-to-tolerance (%P/T). — - Process Capability (Cpk) — Cp and Cpk with specification limits, mean, standard deviation and histogram, by reference designator or part number. Exports to Excel. — - First Pass Yield — Whole-board and sub-panel FPY. Per-position yield in a panel shows where a stencil or reflow problem sits.
- Data & traceability — 2 CAPABILITIES — - MES Export — CSV results, defect crop images, and the whole-board image per unit. Defect codes mapped to your MES vendor codes. — - Inspection History & Reports — Every inspection is stored with its images and verdict, so any board can be traced back. Reports and all analysis views export to Excel.
- Operations & support — 3 CAPABILITIES — - Repair Station — A dedicated screen for the repair operator: the defect call, its location on the board, and the image behind it. — - Offline Programming — Program a product on a machine away from the line, export it, and import it on the production machine. — - Remote Support — Remote assistance from our engineers when a problem needs a second look.

## Industries
- [AOI by Industry](https://daoai.com/industries): How DaoAI AI AOI fits aerospace and defense, medical device, automotive electronics and EMS production.
How DaoAI AI AOI fits aerospace and defense, medical device, automotive electronics and EMS production.

- [AI AOI for Aerospace & Defense PCBA](https://daoai.com/industries/aerospace-defense): On-premise AI AOI for aerospace and defense PCB assembly. All AI runs on the machine, and it programs from one golden board in minutes with no CAD.
AI AOI for Aerospace and Defense PCB Assembly
DaoAI AOI suits aerospace and defense PCB assembly because all AI runs on the machine and inspection data never leaves your factory: no cloud, no account and no internet connection are required, and model training runs offline on the machine. Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes. No CAD, Gerber or component library is required, which fits low-volume, high-mix Class 3 work. DaoAI does not claim any aerospace or defense certification; your quality and export-compliance teams decide how the machine fits your program.
DaoAI AOI suits aerospace and defense PCB assembly because all AI runs on the machine and inspection data never leaves your factory: no cloud, no account and no internet connection are required, and model training runs offline on the machine.
- What makes aerospace and defense PCBA inspection different
- **Class 3 workmanship.** IPC describes Class 3 as "High Performance/Harsh Environment Electronic Products," which "includes products where continued high performance or performance-on-demand is critical, equipment downtime cannot be tolerated." A condition that is acceptable on a Class 2 product can be a defect on a Class 3 product, so the inspection program has to be tight, and every call still gets a human decision. When most of those calls are false, reviewers spend their attention on non-defects.
- **Low volume, many part numbers.** A defense or space program may build a few dozen assemblies, then return to the same design months later. Lines run many short jobs, and each one needs an inspection program before the first board ships. If programming takes hours of engineering time, it competes with the work itself.
- **Board images can be controlled data.** Under the ITAR, "technical data" includes information "required for the design, development, production, manufacture, assembly … of defense articles," including "blueprints, drawings, photographs, plans." Releasing technical data to a foreign person, even inside the United States, is treated as an export. Whether a specific AOI image is ITAR technical data is a determination for your export-compliance team, but many teams start from the assumption that it might be.
- **Cybersecurity requirements in the supply chain.** The DFARS rule that brings the Cybersecurity Maturity Model Certification (CMMC) into defense contracts took effect on November 10, 2025, with a phased rollout. It applies to contractors and subcontractors that process, store or transmit Federal Contract Information or Controlled Unclassified Information. Any shop-floor system that holds program data becomes part of that conversation.
- **Counterfeit-part prevention.** AS9100D clause 8.1.4 requires organizations to "plan, implement, and control processes … for the prevention of counterfeit or suspect counterfeit part use." Inspection is one of the places where wrong or suspect parts can be caught before they ship.
- How DaoAI fits these requirements
- **Inspection data stays in your plant.** All AI runs on the machine; inspection data never leaves your factory. Board images stay on the machine and in the systems you export them to, such as your MES. Model training runs offline, with no image upload. For network-isolated lines, see [On-premise AI AOI and data security](/en-us/daoaiblog/on-premise-ai-aoi-data-security). — **No CAD on the inspection station.** Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes, with no CAD, Gerber or component library required ([how golden-board programming works](/en-us/daoaiblog/aoi-programming-without-cad-files-how-golden-board-programming-works)). — **Controlled changes.** Permissions are set per level: program, tune and train, or inspect only. Locked parameters stay at the values you set, and for each check you can limit the range of AI updates or turn AI updates off ([full control](/features/full-control)). — **Board-level records.** Every inspection is stored with its images and verdict. You can export CSV results, defect crops and the whole-board image for each unit, with defect codes mapped to your MES vendor codes ([production integration](/features/in-your-plant)). — **Marking checks.** [OCR and OCV](/features/ocr) read and verify component markings, date codes and lot codes. Marking verification supports a counterfeit-prevention process but does not replace one. — **Fewer false calls to review.** With tuning, the false-call rate can reach about 1%.
- Recommended DaoAI configurations
- Your situation — Suggested machines — Prototype, NPI and low-volume builds — Production SMT line where component height matters (lifted leads, coplanarity) — Mixed-technology boards with through-hole connectors
- Compare the full range on the [machines page](/machines). For defect-specific guidance, see [lifted lead detection](/en-us/daoaiblog/lifted-lead-defect) and [through-hole AOI](/en-us/daoaiblog/through-hole-tht-aoi-inspection).
- What DaoAI does and does not claim
- DaoAI does not claim ITAR registration, CMMC certification, AS9100 certification or any other aerospace or defense compliance determination. These obligations belong to your organization and your contracts. What we can do is give your quality, IT and export-compliance teams the technical details they need: where data is stored, what the machine connects to, and who can change a program. Sales and engineering support come from DaoAI's North American team in Vancouver, plus authorized partners.
- Send us one of your own boards and we'll test it free of charge.
- Does DaoAI AOI need an internet or cloud connection? — No. All AI runs on the machine, and inspection data never leaves your factory. No cloud, account or internet connection is required, and model training runs offline on the machine. Remote support is optional and off unless you enable it.
- Can DaoAI AOI be programmed without CAD data? — Yes. Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes. No CAD, Gerber or component library is required.
- Is DaoAI ITAR-compliant or CMMC-certified? — DaoAI does not claim either. ITAR and CMMC obligations apply to your organization and contracts. The machine keeps all AI and inspection data on-premise, and your export-compliance team decides how that fits your program.
- Which DaoAI machine fits Class 3 assemblies? — For low-volume and NPI work, the offline P1 or P2. For a production SMT line, the inline 3D P5, which measures component height. For through-hole boards, the A3, which inspects before and after wave solder in one machine.
- Can we stop the AI from changing an approved inspection program? — Yes. Locked parameters stay fixed, and for each check you can limit the range of AI updates or turn them off. Permissions separate who can program, tune and train from who can only inspect.
- IPC, IPC Standards Guidance for Printed Board Assemblies, May 2026 (class definitions)
- ANSI Blog, Acceptability of Electronic Assemblies (IPC-A-610J-2024)
- 22 CFR 120.33, Technical data
- 22 CFR 120.50, Export
- I.S. Partners, CMMC 48 CFR Final Rule Takes Effect November 10, 2025
- ERAI, Latest AS9100 standard brings counterfeit part control to aerospace manufacturing


- [AI AOI for Medical Device PCB Assembly](https://daoai.com/industries/medical-devices): AI AOI for medical device PCBA: on-machine AI, lockable inspection programs, per-unit image records, and golden-board programming with no CAD needed.
AI AOI for Medical Device PCB Assembly
For medical device PCB assembly, DaoAI AOI keeps inspection data on-premise, because all AI runs on the machine. It also lets your quality team decide when the AI may change a program: parameters can be locked, and AI updates can be limited or turned off for each check. Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes, with no CAD, Gerber or component library required. DaoAI does not claim ISO 13485 certification or any FDA compliance determination for its machines.
For medical device PCB assembly, DaoAI AOI keeps inspection data on-premise, because all AI runs on the machine.
- What makes medical device PCBA inspection different
- **The regulation changed in 2026.** FDA's Quality Management System Regulation (QMSR) took effect on February 2, 2026. It incorporates ISO 13485:2016 by reference into 21 CFR Part 820. According to the final rule, its scope is unchanged, which means it still does not extend to manufacturers of components and parts. In practice, requirements reach a PCBA line through the device maker's supplier controls and quality agreements, and some contract manufacturers hold their own ISO 13485 certification.
- **Software that makes quality decisions must be validated.** ISO 13485:2016 clause 4.1.6 requires that software used in the quality management system "be validated prior to initial use and, as appropriate, after changes to such software or its application," with an approach "proportionate to the risk." An AI inspection system that keeps learning raises an obvious question from quality: when does the program change, and who approved it?
- **Records have to tie back to the unit.** Device makers work under UDI (21 CFR Part 830) and, for some devices, tracking requirements (21 CFR Part 821). At the board level, that usually means being able to show what inspection a specific serial number went through and what was found.
- **Acceptance class depends on the product.** Depending on the device and the customer's drawing, boards are built to IPC-A-610 Class 2 or Class 3. IPC describes Class 3 as products "where continued high performance or performance-on-demand is critical, equipment downtime cannot be tolerated."
- **High mix, frequent revisions.** Medical PCBAs are often built in small lots, with engineering changes that each require an updated inspection program. Long programming cycles slow down every revision.
- **Confidential designs.** Device makers protect their board designs closely, and many quality agreements restrict where design data and images can go.
- How DaoAI fits these requirements
- **Change control over the AI.** Locked parameters keep the values you set. For each check you can define the range available to AI updates, or disable AI updates entirely. Auto-programming and model updates keep your manual settings. Permissions separate who can program, tune and train from who can only inspect ([full control](/features/full-control)). — **Inspection records per unit.** Every inspection is stored with its images and verdict. Exports include CSV results, defect crops and the whole-board image for each unit, with defect codes mapped to your MES vendor codes ([production integration](/features/in-your-plant)). [OCR](/features/ocr) reads serial numbers, barcodes and 2D codes on the board. — **Measurement-system data.** Built-in tools run a repeatability study on placement measurements (X offset, Y offset and rotation) against AIAG guidelines and export Cp and Cpk to Excel. Attribute agreement analysis on pass/fail calls is not built in. See [AOI Gauge R&R and Cpk](/en-us/daoaiblog/aoi-gauge-rr-cpk). — **Data stays in your plant.** All AI runs on the machine; inspection data never leaves your factory. Model training runs offline, with no image upload ([data security](/en-us/daoaiblog/on-premise-ai-aoi-data-security)). — **Fast revisions.** Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes, which keeps small lots and engineering changes moving ([AOI for NPI and prototype lines](/en-us/daoaiblog/aoi-for-npi-prototype-lines)). — **Fewer false calls.** With tuning, the false-call rate can reach about 1%.
- Recommended DaoAI configurations
- Your situation — Suggested machines — Small lots, NPI, lab and engineering builds — Production SMT line — Through-hole content before and after wave solder
- Compare all models on the [machines page](/machines).
- What DaoAI does and does not claim
- DaoAI does not claim ISO 13485 certification, FDA registration or any medical-device compliance determination. Validating an AOI for its intended use is the manufacturer's responsibility. We can give your quality team what it needs for that work: how programs are locked, what data the machine stores and exports, and how AI updates are controlled. Sales and engineering support come from DaoAI's North American team in Vancouver, plus authorized partners.
- Send us one of your own boards and we'll test it free of charge.
- Is DaoAI AOI ISO 13485 certified or FDA-cleared? — DaoAI does not claim either. An AOI is production equipment, and the manufacturer validates it for its intended use. DaoAI provides the technical information that validation needs.
- Can we stop the AI from changing a validated inspection program? — Yes. Locked parameters stay at the values you set, and for each check you can limit the range of AI updates or disable them. Permissions control who can program, tune and train.
- Do board images or inspection data leave our facility? — No. All AI runs on the machine; inspection data never leaves your factory. Model training runs offline on the machine. Remote support is optional and off unless you enable it.
- What inspection records can DaoAI export? — Every inspection is stored with its images and verdict. You can export CSV results, defect crops and the whole-board image per unit, with defect codes mapped to your MES vendor codes.
- Does DaoAI support Gauge R&R for AOI? — Yes, for placement measurements: X offset, Y offset and rotation. The built-in study covers repeatability only, with no operator-to-operator comparison; it reports %P/T against the AIAG guideline and exports Cp/Cpk to Excel. Attribute agreement analysis on pass/fail calls is not built in.
- Covington, FDA Finalizes Rule Incorporating ISO 13485 into New Quality Management System Regulation (QMSR), Feb 2024
- FDA, Quality Management System Regulation (QMSR)
- 21 CFR 820.1, Scope
- Greenlight Guru, Software validation under ISO 13485:2016 (clause 4.1.6 text)
- IPC, IPC Standards Guidance for Printed Board Assemblies, May 2026


- [AI AOI for Automotive Electronics PCBA](https://daoai.com/industries/automotive-electronics): Inline 3D AI AOI for automotive electronics: repeatability and Cpk tools, per-unit inspection records for MES, on-machine AI and golden-board programming.
AI AOI for Automotive Electronics PCB Assembly
DaoAI AOI supports automotive electronics lines with inline 3D inspection (P5 and P5D, available now), built-in repeatability and Cpk tools for placement measurements, and per-unit inspection records that export to your MES. Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes, with no CAD, Gerber or component library required. All AI runs on the machine. DaoAI does not claim IATF 16949 certification or any automotive compliance determination.
DaoAI AOI supports automotive electronics lines with inline 3D inspection (P5 and P5D, available now), built-in repeatability and Cpk tools for placement measurements, and per-unit inspection records that export to your MES.
- What makes automotive PCBA inspection different
- **Every inspection system on the control plan needs MSA.** IATF 16949:2016 clause 7.1.5.1.1 states: "Statistical studies shall be conducted to analyse the variation present in the results of each type of inspection, measurement, and test equipment system identified in the control plan." If AOI is on your control plan, it is in scope. Measurement systems analysis is among the five most-cited IATF 16949 nonconformances in IAOB data reported by Smithers in December 2025.
- **Traceability is planned, not improvised.** Summaries of IATF 16949 clause 8.5.2.1 describe an analysis of internal, customer and regulatory traceability requirements, a traceability plan based on risk, and serialization where the customer specifies it. When a field issue comes back, you need to show what inspection a specific unit went through.
- **Volume and variants together.** Automotive lines often run high volumes, but a single module family can have many variants and revisions. The line needs both throughput and fast changeovers.
- **False calls create escape risk.** When operators review long lists of false calls every shift, attention drops, and a real defect is more likely to be passed. Keeping false calls low is a quality measure, not only a cost measure.
- **Height-dependent defects.** Lifted leads, coplanarity problems and tombstoning are easier to judge when component height is measured rather than inferred from a 2D image.
- **Mixed technology.** Connectors and power components are often through-hole and wave-soldered, so inspection is needed after wave as well as after reflow.
- How DaoAI fits these requirements
- **Inline 3D, available now.** [P5](/machines/p5) (single track) and [P5D](/machines/p5d) (dual track) measure component height, with height accuracy ≤ 6 µm with the calibration tool (reference value). For 2D at higher throughput, [P3D](/machines/p3d) is dual-track inline. See [offline vs inline vs 3D AOI](/en-us/daoaiblog/offline-vs-inline-vs-3d-aoi). — **MSA support.** Built-in tools run a repeatability study on placement measurements (X offset, Y offset, rotation) against AIAG guidelines and export Cp and Cpk to Excel. Attribute agreement analysis on pass/fail calls is not built in, so plan that study separately. See [AOI Gauge R&R and Cpk](/en-us/daoaiblog/aoi-gauge-rr-cpk). — **Unit-level records.** Every inspection is stored with its images and verdict. Exports include CSV results, defect crops and the whole-board image per unit, with defect codes mapped to your MES vendor codes ([production integration](/features/in-your-plant)). [OCR](/features/ocr) reads serial numbers, barcodes and 2D codes. — **Low false calls.** With tuning, the false-call rate can reach about 1% ([how to reduce AOI false calls](/en-us/daoaiblog/how-to-reduce-aoi-false-calls-root-causes-and-fixes)). — **Stable programs.** Locked parameters stay fixed, and AI updates can be limited or turned off per check, so a released program does not drift ([full control](/features/full-control)). — **Fast variant changeovers.** Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes. — **Data stays in your plant.** All AI runs on the machine; inspection data never leaves your factory.
- Recommended DaoAI configurations
- Your situation — Suggested machines — High-volume SMT line, height-critical parts — High-throughput line where 2D is enough — Through-hole connectors and power parts — Samples, NPI and engineering builds
- Compare all models on the [machines page](/machines).
- How to evaluate an AOI for an automotive line
- Bring a golden board plus boards with known defects from your own process, including height-related ones. — Time the programming of each variant, from first board to running inspection. — Run the repeatability study on placement measurements you would put on the control plan, and plan the attribute study separately. — Count false calls per board over a realistic run, not a demo panel. — Check that the exported records match the fields your MES and traceability plan expect.
- What DaoAI does and does not claim
- DaoAI does not claim IATF 16949 certification or any automotive compliance determination. IATF 16949 applies to your quality management system, and your team decides how the AOI fits your control plan and MSA program. We provide the data and tools that work needs. Sales and engineering support come from DaoAI's North American team in Vancouver, plus authorized partners.
- Send us one of your own boards and we'll test it free of charge.
- Is DaoAI IATF 16949 certified? — DaoAI does not claim IATF 16949 certification. IATF 16949 applies to the automotive supplier's quality management system; DaoAI provides MSA tools and inspection records your team can use within it.
- Can DaoAI AOI support a Gauge R&R study? — Yes, for placement measurements: X offset, Y offset and rotation. The built-in study covers repeatability only, with no operator-to-operator comparison; it reports %P/T against the AIAG guideline and exports Cp/Cpk to Excel. Attribute agreement analysis on pass/fail calls is not built in.
- Does DaoAI offer inline 3D AOI? — Yes. The P5 (single track) and P5D (dual track) are inline 3D AOI machines and are available now.
- How does DaoAI support traceability? — Every inspection is stored with its images and verdict. You can export CSV results, defect crops and the whole-board image per unit, with defect codes mapped to your MES vendor codes.
- Can one machine inspect through-hole connectors before and after wave solder? — Yes. The A3 inspects through-hole boards before and after wave solder with one machine.
- ASQ discussion quoting IATF 16949:2016 clause 7.1.5.1.1
- Smithers, Navigating IATF 16949: Top 5 Non-Conformances via the IAOB, Dec 2025
- 16949store, Identification and traceability (IATF 16949 clause 8.5.2.1 summary)


- [AI AOI for EMS & Contract Manufacturers](https://daoai.com/industries/ems-contract-manufacturers): AI AOI for high-mix EMS lines: program each new customer job from one golden board in minutes, no CAD needed. Customer data stays in your factory.
AI AOI for EMS and Contract Manufacturers
DaoAI AOI is built for high-mix EMS and contract manufacturing lines. Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes. No CAD, Gerber or component library is required, so a new customer job goes from new board to running inspection in minutes, not a shift. All AI runs on the machine, so customer board images and inspection data never leave your factory. The range covers offline 2D (P1, P1L, P2), inline 2D (P3, P3D), inline 3D (P5, P5D, available now) and through-hole (A3, pre- and post-wave in one machine).
DaoAI AOI is built for high-mix EMS and contract manufacturing lines.
- What makes EMS inspection different
- **Many customers, many short jobs.** A contract manufacturer may run several new part numbers a week, often on the same line. Each job needs an inspection program before the first board ships, and programming time multiplies with every changeover. When one engineer owns AOI programming, that person becomes the line's bottleneck.
- **CAD data is often missing or late.** Customers send kits without usable CAD, centroid files arrive incomplete, or legacy boards have no data at all. An AOI that needs CAD and a component library to start can't inspect the job until someone rebuilds that data.
- **Each customer sets its own acceptance class.** IPC-A-610 defines three classes: Class 1 for "General Electronic Products," Class 2 for "Dedicated Service Electronic Products," and Class 3 for "High Performance/Harsh Environment Electronic Products." The same line may build all three in one week, with different thresholds for each job.
- **Customer data is not yours to share.** Board images show your customers' designs. NDAs and quality agreements often limit where that data can go, and customers in defense, medical and automotive markets ask more questions about it.
- **Customers ask for traceability.** IPC-1782B, the IPC standard for manufacturing and supply chain traceability, sets "minimum requirements for manufacturing and supply chain traceability based on perceived risk." More customers expect board-level inspection records.
- **Margins depend on operator time.** False calls cost review time on every board, and long programming cycles keep a line waiting.
- How DaoAI fits EMS operations
- **Golden-board programming.** Every DaoAI AOI programs itself from one golden board in 30 seconds to 5 minutes, with no CAD, Gerber or component library ([auto-programming](/features/auto-programming)). CAD is optional, never required. — **Program away from the line.** You can program a product on another machine, export it, and import it on the production machine, so the SMT line keeps running ([production integration](/features/in-your-plant)). — **Panels and arrays.** See [panel inspection](/features/panel) and [panel and array AOI](/en-us/daoaiblog/panel-array-aoi-inspection). — **Per-job control.** Thresholds are adjustable per check, and permissions separate who can program, tune and train from who can only inspect ([full control](/features/full-control)). — **Customer data stays in your plant.** All AI runs on the machine; inspection data never leaves your factory ([data security](/en-us/daoaiblog/on-premise-ai-aoi-data-security)). — **Records customers can audit.** Every inspection is stored with its images and verdict. Exports include CSV results, defect crops and the whole-board image per unit, with defect codes mapped to your MES vendor codes. — **Low false calls.** With tuning, the false-call rate can reach about 1%. — **Support where you are.** Sales and engineering support come from DaoAI's North American team in Vancouver, plus authorized partners. There are 1,000+ systems in the field.
- For a step-by-step evaluation framework, see [how to choose an AOI for high-mix PCB assembly](/en-us/daoaiblog/how-to-choose-aoi-for-high-mix-pcb-assembly) and [how to calculate AOI ROI and payback](/en-us/daoaiblog/how-to-calculate-aoi-roi-and-payback-period).
- Recommended DaoAI configurations
- Your shop — Suggested machines — High-mix, low-volume or NPI-heavy, no conveyor needed — One or two SMT lines — Dual-lane, higher-throughput lines — Mixed technology with wave solder — Add the [A3](/machines/a3), pre- and post-wave in one machine
- Compare all models on the [machines page](/machines).
- How to run a fair trial on your own jobs
- Pick two or three jobs that are hard to program today: dense boards, missing CAD, or frequent revisions. — Time each one from golden board to running inspection. — Count false calls per board over a full lot, and note how many your operators would have to review each shift. — Check that the exported records cover what your customers ask for in their quality agreements. — Ask who supports the machine in your time zone, and how fast.
- Send us one of your own boards and we'll test it free of charge. Pick a job that is hard to program today.
- Do we need CAD data from our customers to program DaoAI AOI? — No. Every DaoAI AOI programs itself from one golden board. No CAD, Gerber or component library is required. CAD is optional, never required.
- How long does it take to program a new job? — Program generation takes 30 seconds to 5 minutes from one golden board. DaoAI reports that a new board is typically ready for production inspection in about 5 to 10 minutes.
- Can we program a job without stopping the SMT line? — Yes. You can program a product on another machine, export the program, and import it on the production machine.
- Will our customers' board images leave our factory? — No. All AI runs on the machine; inspection data never leaves your factory.
- Can we test DaoAI on our own boards before buying? — Yes. Send us one of your own boards and we'll test it free of charge.
- IPC, IPC Standards Guidance for Printed Board Assemblies, May 2026
- DIN Media, IPC-1782B: Standard for Manufacturing and Supply Chain Traceability of Electronic Products, 2023-09-01

## Key guides
- [How to Calculate AOI ROI and Payback Period](https://daoai.com/en-us/daoaiblog/how-to-calculate-aoi-roi-and-payback-period): AOI ROI without vendor math: the cost lines to count, formulas for false-call and escape savings, a worked payback example, and when it does not pay.
- [How to Evaluate an AI AOI System Before You Buy](https://daoai.com/en-us/daoaiblog/how-to-evaluate-an-ai-aoi-system-before-you-buy): Vendor accuracy numbers are marketing until you reproduce them. A pilot protocol for AOI evaluation: which boards to bring, what to measure, how to score.
- [How to Choose an AOI for High-Mix PCB Assembly](https://daoai.com/en-us/daoaiblog/how-to-choose-aoi-for-high-mix-pcb-assembly): Choosing an AOI for high-mix PCB assembly? Specs are table stakes. The real differences: programming inputs, changeover time, and false-call burden.
- [How to Reduce AOI False Calls: Root Causes & Fixes](https://daoai.com/en-us/daoaiblog/how-to-reduce-aoi-false-calls-root-causes-and-fixes): Why does your AOI throw so many false calls? The root causes, from same-color parts to marking interference, and the fixes that work, traditional and AI.
- [AOI Programming Without CAD Files: How It Actually Works](https://daoai.com/en-us/daoaiblog/aoi-programming-without-cad-files-how-golden-board-programming-works): How does AOI programming without CAD files work? A step-by-step look at golden-board, feature-cognition programming vs the traditional 3–5 hour CAD workflow.
- [How to Cut AOI Changeover Time on High Mix SMT Lines](https://daoai.com/en-us/daoaiblog/how-to-cut-aoi-changeover-time-on-high-mix-smt-lines): Cut AOI changeover time with job preparation, revision checks and a clear first article release process for high mix SMT production.
- [AOI for NPI and Prototypes: Inspect the First Board, Not the 100th](https://daoai.com/en-us/daoaiblog/aoi-for-npi-prototype-lines): Why NPI and prototype builds skip AOI, how golden-board programming inspects the first board of a build, and where offline AOI beside the line fits.
- [Gauge R&R and Cpk for AOI: Proving Your Inspection Is Capable](https://daoai.com/en-us/daoaiblog/aoi-gauge-rr-cpk): What Gauge R&R and Cpk mean for an AOI system, how to run a GR&R study step by step, which results are acceptable, and what to control during the study.
- [How Much Does an AOI Machine Cost? What Drives the Price](https://daoai.com/en-us/daoaiblog/aoi-machine-cost-factors): What sets the price of an AOI machine: 2D vs 3D, inline vs offline, dual lane, camera, software licensing, plus programming, false-call and service costs.
- [AOI Metrics: False Call Rate, Escape Rate and First Pass Yield](https://daoai.com/en-us/daoaiblog/aoi-performance-metrics): How to define and measure AOI false call rate, escape rate and first pass yield on your SMT line, with plain-text formulas and why loose thresholds cost you.
- [AOI Specifications Explained: Camera, Resolution, FOV, Lens and Speed](https://daoai.com/en-us/daoaiblog/aoi-specifications-explained): How to read an AOI spec sheet: megapixels, µm per pixel, field of view, telecentric lenses, lighting, FOV time, board limits and what matters for 01005/0201.
- [AOI vs SPI vs AXI: Which Inspection Belongs Where in the SMT Line](https://daoai.com/en-us/daoaiblog/aoi-vs-spi-vs-axi): AOI vs SPI vs AXI: what each inspection sees and misses, where it sits in the SMT line (post-print, pre-reflow, post-reflow, post-wave) and when you need it.
- [Offline vs Inline vs 3D AOI: How to Choose the Right Machine](https://daoai.com/en-us/daoaiblog/offline-vs-inline-vs-3d-aoi): Offline, inline 2D or 3D AOI? Choose by volume, product mix, board size, height-sensitive defects and budget, with real specs from the DaoAI AOI range.
- [On-Premise AI AOI: Why Board Images Should Stay in Your Plant](https://daoai.com/en-us/daoaiblog/on-premise-ai-aoi-data-security): Why AOI board images are customer IP, how NDAs and ITAR-style rules apply, and how offline AI training lets AOI run on network-isolated lines in your plant.
- [Panel and Array AOI: Program One Board, Inspect the Whole Panel](https://daoai.com/en-us/daoaiblog/panel-array-aoi-inspection): How panel (array) AOI inspection works: program one sub-panel, align every position, handle fiducials and bad marks, and keep each board traceable.
- [OCR and OCV on PCBs: Verifying Component Markings with AI](https://daoai.com/en-us/daoaiblog/pcb-ocr-ocv-component-marking): How AOI uses OCR and OCV to verify PCB component markings: catching wrong parts, flagging suspect markings, and reading serials and barcodes for traceability.
- [Through-Hole (THT) AOI: Pre-Wave and Post-Wave Inspection Guide](https://daoai.com/en-us/daoaiblog/through-hole-tht-aoi-inspection): How through-hole AOI works: THT defects, what pre-wave and post-wave checks catch, why manual inspection misses them, and how to add AOI to a wave line.
- [What Is AI AOI? How AI-Powered Automated Optical Inspection Works](https://daoai.com/en-us/daoaiblog/what-is-ai-aoi): AI AOI uses trained models, not hand-built rules, to inspect PCBAs. How rule-based AOI works, what AI changes, 2D vs 3D, and what to ask vendors.
- [Koh Young AOI Alternative: DaoAI vs Koh Young Compared](https://daoai.com/en-us/daoaiblog/koh-young-aoi-alternative): DaoAI vs Koh Young AOI compared: golden-board vs Gerber-based programming, inline 3D specs, AI false-call handling, through-hole, on-premise data and cost.
- [Mirtec AOI Alternative: DaoAI vs Mirtec Compared](https://daoai.com/en-us/daoaiblog/mirtec-aoi-alternative): DaoAI vs MIRTEC AOI compared: golden-board AI programming vs library-based auto teaching, 3D specs, AI false-call handling, THT, data and cost of ownership.
- [Omron AOI Alternative: DaoAI vs Omron VT Series](https://daoai.com/en-us/daoaiblog/omron-aoi-alternative): Comparing Omron VT-S 3D AOI with DaoAI: programming inputs, 3D specs, AI features, offline and THT options, data handling, and how to test both on your boards.
- [Saki AOI Alternative: DaoAI vs Saki 3D AOI](https://daoai.com/en-us/daoaiblog/saki-aoi-alternative): Comparing Saki 3Di-EX 3D AOI with DaoAI: CAD-based vs golden-board programming, 3D range, board size, AI, THT options, and how to test both on your boards.
- [Viscom AOI Alternative: DaoAI vs Viscom](https://daoai.com/en-us/daoaiblog/viscom-aoi-alternative): Comparing Viscom iS6059 and vAI with DaoAI: E-CAD vs golden-board programming, 3D metrics, AI verification, offline and THT options, and how to test both.
- [YESTECH AOI Alternative: DaoAI vs Nordson YESTECH Compared](https://daoai.com/en-us/daoaiblog/yestech-aoi-alternative): DaoAI vs Nordson YESTECH AOI: golden-board vs CAD-and-library programming, inline 3D, through-hole, on-machine AI, and how to test both on your boards.
- [CyberOptics AOI Alternative: DaoAI vs Nordson SQ-Series 3D AOI](https://daoai.com/en-us/daoaiblog/cyberoptics-aoi-alternative): DaoAI vs CyberOptics (Nordson) SQ-series 3D AOI: golden-board vs library-based programming, inline 3D, through-hole, on-machine AI and how to test both.
- [ViTrox AOI Alternative: DaoAI vs ViTrox V510 3D AOI](https://daoai.com/en-us/daoaiblog/vitrox-aoi-alternative): DaoAI vs ViTrox V510 3D AOI: how each handles AI programming without CAD, inline 3D, through-hole after wave, where the AI runs, and how to test both.
- [TRI AOI Alternative: DaoAI vs TRI TR7700 Series 3D AOI](https://daoai.com/en-us/daoaiblog/tri-aoi-alternative): DaoAI vs TRI (Test Research, Inc.) TR7700 3D AOI: golden-board programming, inline 3D, through-hole after wave, on-machine AI and a test plan for both.
- [Mek AOI Alternative: DaoAI vs Mek (Marantz Electronics) Compared](https://daoai.com/en-us/daoaiblog/mek-aoi-alternative): DaoAI vs Mek (Marantz Electronics) AOI: golden-board vs EZPro AI-assisted programming, inline and desktop 3D, through-hole, and how to test both.
- [Used AOI Machine Buying Checklist for High Mix PCB Assembly](https://daoai.com/en-us/daoaiblog/used-aoi-machine-buying-checklist): Buying a used AOI machine? Check software, condition, product changes and the cost of getting into production with this practical PCB inspection checklist.
- [Become a DaoAI Distributor: Bring AI-Powered AOI to High-Mix, Low-Volume Manufacturing](https://daoai.com/en-us/daoaiblog/become-a-daoai-distributor-ai-aoi-high-mix-low-volume): DaoAI is recruiting SMT and electronics-manufacturing equipment distributors. See why high-mix, low-volume factories need AI AOI, what partners get, and how to apply.

## Company
- [About DaoAI: AI AOI Manufacturer in Vancouver](https://daoai.com/about): About DaoAI, an AI AOI manufacturer founded in 2018 in Vancouver, British Columbia, building AI-native automated optical inspection for PCB and PCBA lines.
- [Become a DaoAI AOI Distributor or Sales Rep | North America](https://daoai.com/distributors): Recruiting SMT equipment distributors and commission sales reps in the US, Canada and Mexico. AI-native AOI lineup, North American engineering support, training and demo machines.
- [AOI Brochure, Spec Sheet and Whitepaper](https://daoai.com/downloads): Download the DaoAI AOI brochure, P-Series spec sheet and NPI whitepaper: three documents on AI auto-programming, machine specs and inspection, for engineers.
- [Meet DaoAI at SMTA International 2026 — Booth 2727](https://daoai.com/events/smta-international-2026): Meet DaoAI at SMTA International 2026, Booth 2727, Oct 27–29 in Rosemont, IL. Talk AI auto-programming, inline 3D AOI, THT inspection and partnering.
- [Blog](https://daoai.com/blog): AOI and SMT inspection blog by the engineers who build DaoAI machines: AOI programming, false calls, high-mix changeover, ROI and PCBA defect detection.
- Contact: business@daoai.com (sales and distribution)