Skip to main content

navio VISION

navio VISION is a visual anomaly detection solution for automated quality inspection, built as an extension on top of the navio platform. It lets you train a model that distinguishes normal from anomalous images of your parts or production states — without any machine learning expertise — and use it to inspect images manually or directly from your production line.

navio VISION use cases

How It Relates to navio​

navio VISION provides its own, simplified user interface tailored to visual inspection. Under the hood, everything is powered by the navio platform:

  • Every VISION use case is a navio use case.
  • Model training, deployments and inference run on navio.
  • Image labeling uses navio's data labeling integration.
  • Retraining with labeled production data is handled by navio's retraining pipeline.

You don't need to know any of this to use navio VISION — but if you want to dig deeper, the linked platform documentation applies.

The Workflow​

  1. Create a use case — provide images of normal (OK) and anomalous (not OK) states; a model is trained and deployed automatically.
  2. Inspect images — upload images and get an instant Normal/Anomaly classification.
  3. Connect live data — ingest images continuously from your production line (S3 or Azure Blob Storage) for automatic analysis.
  4. Label and retrain — confirm or correct the model's predictions and retrain to improve performance.
  5. Monitor results — track normal/anomaly fractions over time and browse the inspection history.

Availability​

navio VISION is available as SaaS at vision.navio.ai. A self-hosted installation alongside your own navio instance is available on request — contact us if you are interested.