Solutions

Plant data and Connector Hub

Connect site or enterprise databases, or start from uploaded files. Canonical mappings are the path for data that has to arrive every month.

Request a quote Why teams choose AEYS

ISO 50001:2018

Energy models are only as trustworthy as the columns behind them. If every training run reopens the question of which field is the timestamp, which is consumption, and which are relevant variables, the EnMS never settles.

Connector Hub connects site or enterprise databases. Canonical mappings bind timestamp, consumption, and relevant variables per unit and energy carrier so training and predict do not re-ask which column is the clock. Uploaded files still work for the first model; the hub is the path for data that has to arrive every month.

What you get

Canonical mappings per unit and carrier: Timestamp, consumption, and relevant variables are bound once. Models inherit the mapping instead of a new spreadsheet contract.
Files first, historian when you are ready: The first SEU models can be trained from uploads. Connector Hub takes over when the same extract must arrive on a schedule.
Site or enterprise databases: The hub is built for the customer’s own Postgres (or similar) — AEYS does not need to become the plant historian.
Predict without re-mapping: Once the clock column is agreed, expected-versus-actual runs reuse it. Operators are not asked to label columns again.

Frequently asked

Do we need a historian on day one?: No. Models can be trained from uploaded files. Connector Hub and canonical mappings are the path for recurring operational data once the first SEUs are agreed.
Does AEYS store a second copy of every meter?: No. The hub maps the customer database into the training and predict path. AEYS is not a replacement plant historian.

What you get

Canonical mappings per unit and carrier

Timestamp, consumption, and relevant variables are bound once. Models inherit the mapping instead of a new spreadsheet contract.

Files first, historian when you are ready

The first SEU models can be trained from uploads. Connector Hub takes over when the same extract must arrive on a schedule.

Site or enterprise databases

The hub is built for the customer’s own Postgres (or similar) — AEYS does not need to become the plant historian.

Predict without re-mapping

Once the clock column is agreed, expected-versus-actual runs reuse it. Operators are not asked to label columns again.

Frequently asked

Do we need a historian on day one?

No. Models can be trained from uploaded files. Connector Hub and canonical mappings are the path for recurring operational data once the first SEUs are agreed.

Does AEYS store a second copy of every meter?

No. The hub maps the customer database into the training and predict path. AEYS is not a replacement plant historian.