RESOURCES

Download resources.

Brochures, demo videos and whitepapers — leave your details to get the full materials.

DOWNLOADS

FAQ

How many samples to start?

One good (positive) sample is enough to build a model — no defect images, no large datasets. That's the core of DaoAI positive-sample learning.

How long does deployment take?

About 30 seconds to model from one sample and 10 minutes to ramp up; line operators run it themselves, no AI engineer needed.

Which defects can it detect?

Missing parts, tombstone, bridge, polarity, OCR, lifted leads, solder wicking, package lift and 30+ more 2D / 3D defects, covering components and solder joints.

Is the data secure?

100% on-premise — data never leaves the factory and there is no cloud dependency, meeting government and high-compliance requirements.

Can it connect to MES / line systems?

It imports CAD coordinates, auto-programs, supports barcode traceability and MES file integration — fitting seamlessly into existing lines.

Will the model degrade over time?

On the contrary — a feedback loop returns review results to update the model in minutes, so it keeps improving.