Can this run on top of our PI historian without writing to it?
Yes. We design read-only adapters against the historian and keep application data in a separate, client-approved store, so the historian remains the system of record and is never written to by the reporting layer.
How are alerts kept from becoming noise?
Every rule carries a duration, a data-quality and freshness check, a sample-gap tolerance, hysteresis and a cooldown. 'Seen' is separate from 'recovered', so an acknowledged alert does not silently close.
Do you replace our existing SCADA or work with it?
Usually we work with it. Most projects integrate an existing SCADA or historian into a modern supervisory and reporting layer, and only replace components that are unsupported or unsafe. Replacement is a scoped decision, not a default.
How do you keep control networks secure when adding remote monitoring?
Control networks stay segmented. We read from historians or edge gateways through read-only adapters, publish to a separate cloud layer, and never expose PLCs or SCADA servers directly to the internet. Access is role-based and logged.
Can operators and executives use the same system?
Yes, with different views. Operators get HMI screens, alarms and trends; executives get regional and asset KPIs, history and personal alerts on mobile. Both read the same data model, so the numbers agree.
Which protocols and platforms do you integrate?
Modbus TCP/RTU, OPC UA, MQTT, common PLC vendor protocols, and historians such as PI and SQL-based time-series stores. Where a protocol is proprietary we work through the vendor's supported interface rather than reverse-engineering it.