Process data analytics your engineers do themselves, ending in an action.

Search years of historian data by value, line up the periods that matter and check the loops behind them, without a data scientist or a spreadsheet. What your engineers find becomes a tracked action, not a screenshot in an email.

Delivered by Cyzag Operations Cyzag Intelligence

Book a working session → What it does

Value search · last 90 days

FIC-220 flow below 30 m³/h for more than 10 min

4 periods found · 1.4 hours in all

3 Jul · 06:10 42 min
28 Jun · 14:55 18 min
21 Jun · 02:30 12 min
14 Jun · 09:20 12 min

Value-based search: every period FIC-220 flow stayed below 30 m³/h for more than 10 minutes

What it does

Six things it does.

Built with Analysis workspace Batch and loss analysis Parameter drill-down Cause-and-effect flow Historians

Search the history by value

Every period a tag crossed a limit, for as long as you set, across years of historian data.
Trends on one canvas

Charts dragged into a grid, periods lined up side by side, the layout saved as a template for the next unit.
Batches compared step by step

Any batch against your best ones, with downtime and losses overlaid on the trend.
Control-loop checks

Oscillation and saturation found, and a valve problem told from a tuning problem.
The shift's context on the trend

Log entries, deviations and flags shown where they happened, so a dip has its explanation beside it.
From finding to action

A task or a root-cause workflow opened from the chart, with the evidence attached.

How it works

How it works, in four steps.

1 Read-only connections bring in the historian; nothing migrates.
2 Engineers open the workspace on any unit and search the history by value.
3 Periods, batches and loops are compared on one canvas, with the shift's records on the trend.
4 What they find becomes a task or a root-cause workflow, and a named person approves the fix.

Who does what

Cyzag Operations → Shows the shift's log entries and deviations on the same trend, and turns what you find into a tracked action.
Cyzag Intelligence → Puts its flags and the likely cause on the trend you are looking at, with the evidence behind them.

Where it runs

Read-only connections to your historian, MES and SAP
Deployed on site or in your own tenant
Configured by your engineers, with no code
An audit trail on every record

What it connects to →

Questions

What buyers ask us.

Do we need a data scientist?

No. The workspace is built for process engineers: charts, searches and loop checks are set up by pointing and clicking, not by writing code.

What is value-based search?

A search over the historian for every period that meets a condition, such as a flow below 30 m³/h for more than 10 minutes. It shows how often the condition occurs and for how long, and lets you line the periods up to compare them.

How is this different from a trending tool?

A trending tool stops at the chart. Here the shift's records sit on the same trend, the engine proposes the likely cause, and what you find becomes a tracked action that is checked afterwards.

Does it replace our historian?

No. Your historian stays where it is and Cyzag reads from it. AVEVA PI, GE Proficy, Aspen IP.21, Canary, TimescaleDB and InfluxDB are supported.

Can it summarise what has changed?

Yes, with your AI service switched on: an AI summary says what has changed since you last looked.