Own the production evidence.
Your production images and labeled examples preserve what acceptable and defective conditions look like. Use that evidence to validate changes and keep inspections aligned with your products and processes.
Data ownership
Your images. Your trained models. Your inspection IP.
Every inspection can build more than a verdict. You own your production images, trained AI models and customer-specific inspection IP. United Vision manages training in your chosen, customer-owned cloud account and runs production inference on your supported factory-owned edge devices, keeping inspection knowledge a reusable factory asset.
In brief
The customer owns the production evidence, trained AI models and customer-specific inspection IP developed with United Vision. Training and retraining run in a supported cloud account your company chooses and owns, not a vendor-owned SaaS account. Production inference always runs at your factory edge.
Ownership questions
Build inspection knowledge your factory keeps.
Own the models your production teaches.
Your trained AI models, defect definitions and customer-specific inspection IP remain yours. Retained evidence and model versions give the next improvement a starting point, rather than treating each update as a new project.
Keep training in your cloud account.
United Vision manages training and retraining in the supported cloud account your company chooses and owns. Approved models deploy to your supported edge devices for production inference.
Procurement checklist
Six data-control questions, answered.
Permitted use
May production data or customer-specific models be used outside the agreed service?
The agreement should limit customer production data and customer-specific inspection work to delivering, supporting and improving the agreed customer service. United Vision does not use customer IP to train or improve shared, general or third-party AI models. Any additional permitted use should be explicit rather than assumed.
Access
Which customer and provider roles can access images, labels, models and results?
The customer should define who can view evidence, label images, train models, approve releases and review results. Provider support access, if required, should be documented separately with its purpose and limits. The customer-owned cloud account keeps administrative control with the customer while role design remains an implementation decision.
Retention
What is retained, for how long, and how is deletion handled?
Set a policy by data type and environment: raw production images, labeled datasets, model artifacts, inspection results, logs and backups may need different retention periods. The agreement should also define deletion requests, backup expiry and any legal retention requirement. A customer-owned account provides control, but the policy still needs to be explicit.
Portability
What can the customer export if the operating model changes?
You own your trained AI models and customer-specific inspection IP. Document the export formats, access and transition support for images, labels, model versions, configurations, results and logs. Ownership does not by itself guarantee compatibility with another runtime or continued use of United Vision software after termination. The agreement should define those boundaries.
Security
How are environments, credentials, transfers and audit events governed?
Use a documented responsibility matrix for customer identities, service credentials, secrets, network boundaries, data transfers, logging, support access and incident escalation. The exact controls depend on the supported customer cloud and factory-edge configuration, so they should be reviewed against the customer’s standards and recorded in the security schedule.
Responsibility
Which commitments belong in the commercial agreement, privacy notice and security schedule?
The commercial agreement should define service scope, data use, ownership, access, retention, export, deletion, support and exit responsibilities. The privacy notice should address applicable personal-data processing. The security schedule should record the agreed technical and operational controls. Product-page language explains the intended model; executed documents define the binding commitments.
Questions answered
What buyers ask about data ownership.
Who owns our trained AI models, production images and inspection IP?
The customer owns its production images, trained AI models and customer-specific inspection IP developed with United Vision. Labeled evidence, defect definitions and model versions retain production knowledge that can support validation and future updates. This is ownership of the customer-specific inspection assets, not ownership of the United Vision software platform.
What does customer-controlled cloud mean?
It means your company chooses and owns the supported cloud account where United Vision manages model training and retraining. It is not a tenant in a vendor-owned SaaS training environment. The provider, region, roles, security controls and support-access responsibilities are defined for your environment; production inference always runs at the factory edge.
Why does plant ownership matter for defect detection?
Defect definitions and production conditions change. Retaining your images, trained models and inspection IP means the knowledge built through one inspection remains a factory asset. Teams can review evidence and update the model through the same no-code workflow as production changes, while each new use case is qualified against its own requirements.
Where does production inference run?
Production inference always runs on a supported factory-owned edge device close to the line. United Vision manages model training and retraining in your chosen, customer-owned cloud account. The cloud is not the real-time camera-to-verdict path.
Does customer ownership include United Vision's software?
No. Customer ownership covers your production images, trained AI models and customer-specific inspection IP. United Vision retains its software platform and provider intellectual property. Software access, license scope, renewal and termination are governed by the agreement.
Can we export our inspection assets or move them to another system?
Export formats, access, transition support and post-termination responsibilities must be defined in the agreement. Customer ownership does not automatically make a trained model compatible with another runtime or provide perpetual rights to use United Vision software. Confirm the required image, label, model and result exports before deployment.
Bring one inspection opportunity. Leave with a clearer path.
A working session is designed around the image source, operating condition, decision, response and ownership model, not a generic product tour.