FAQ
Common questions about adoption, quality, data, IP and workflow.
Does PLC Automator replace controls engineers?
No. The product prepares candidate data; a qualified engineer reviews exceptions and approves the final output. It is a workflow tool, not a replacement for engineering judgement.
Can the AI make a mistake?
Yes. The model output is treated as candidate data. Uncertain, incomplete or conflicting information is flagged as an exception for engineer review, not accepted as an automatic engineering decision.
What drawing types and systems are supported?
A pilot starts with one agreed drawing type and export. Universal support for every format, vendor or system is not claimed; each workflow is agreed and demonstrated before it is relied on.
What happens with messy, incomplete or ambiguous drawings?
Those items are surfaced as review exceptions. They are not hidden or auto-completed. An engineer inspects the source evidence and decides how to proceed.
Are our drawings confidential?
Yes. Drawings are treated as customer-owned and confidential. Only the material authorised for the agreed workflow is processed, and access is limited to authorised users.
Who owns the drawings, outputs and product configuration?
The customer retains ownership of project drawings, project information and project-specific outputs. Product and company IP remain distinct and are covered by the applicable agreement.
Is customer data used to train AI models?
Customer project data is not used for model training by default. Any exception requires explicit written permission and a processing configuration that supports it.
Does the product access live PLC/SCADA/OT systems?
No. The product processes authorised engineering files. It does not connect to, monitor or deploy changes to live OT systems.
How does this work with our tag naming rules, templates and exports?
We load the customer’s naming rules, templates and export formats before processing. The output is a candidate data pack that matches the agreed rules, demonstrated for the specific workflow.
Does this reduce our billable work?
It reduces repeated data transcription. Under strict time-and-materials contracts, actual labour must still be recorded. The value is in capacity, quality, faster delivery and better review—not in fabricated labour.
Can we export our data and avoid lock-in?
Yes. Project outputs are customer-owned CSV/Excel data packs. You can use them with the downstream workflow without a proprietary dependency.
Can you configure or co-develop a workflow for our company?
Configuration and bespoke co-development can be discussed, but each must be separately scoped and agreed. They are not included in a standard subscription without an explicit statement of work.
What does a controlled pilot involve?
One agreed drawing type, rule set and export. We measure preparation time, exceptions, engineer review effort and output usefulness before proposing broader adoption.
Can we discuss a private or customer-controlled processing route?
Yes. Private AWS processing and customer-controlled environments can be discussed under planned higher-assurance options. The architecture, data residency, access controls, retention and responsibilities must be agreed before deployment.