Skip to content
Hakan KaracaENGINEERING JOURNAL
PLC & SCADAEki 8, 20262 min read

Industrial Data Collection with Ignition SCADA: From Tag to Trusted Record

Ignition SCADA'nın mimari yaklaşımını, tag modelini ve SQL entegrasyonunu anlatan teknik yazı.

Connecting a tag is the easy part; a usable data record needs context, quality and a recovery plan.

Platforms such as Ignition can connect industrial data sources to supervisory views, historians, SQL systems and web-based applications. A dependable collection design starts by stating the purpose of each data point. Is it a live process value, an alarm, an equipment state, a production counter, a batch transition or an operator-entered reason? The purpose drives the sampling strategy, storage model and validation rules.

Organise tags around meaning

Use consistent names, engineering units, data types, source ownership and descriptions. A hierarchy should help an engineer identify plant, line, cell and equipment without relying on a tribal naming convention. Keep vendor-specific addressing separate from the business-facing name where possible, and record the mapping as controlled configuration.

Choose collection intervals deliberately

Collecting every tag at the fastest available rate is rarely a good default. A temperature loop may need a dense time series for process analysis, while a production order state changes only occasionally. Sampling rate affects network load, storage, query performance and the ability to interpret a trend. For event data, a change-driven model can be more meaningful than repeatedly storing identical values.

Separate time-series samples from business events

A historian is suited to time-series measurements. A discrete event table is suited to records with identity and meaning, such as order started, batch completed or quality disposition changed. Those records should generally have stable IDs, source and ingestion timestamps, equipment and order context, plus an explicit event type. Do not force every business event into a single numeric tag and expect the meaning to remain obvious.

Plan for disconnected sources and databases

Define store-and-forward behaviour, buffer capacity and the alarm policy when a destination becomes unavailable. Test what happens when the database server is restarted or a certificate expires. Confirm that data collected during the outage are replayed with the original source time and do not create unexplained duplicates.

Commission with evidence

  • Compare representative source values with the values displayed and stored.
  • Check units, scaling, deadbands, timestamps and quality flags.
  • Simulate a source disconnect and restore.
  • Verify database permissions, retention, backup and restore.
  • Review alarms for actionable description, priority and ownership.
  • Record the software version, gateway settings and change history.

In production, “the dashboard updates” is only a first test. Ask whether the value is the right value, whether it is fresh, whether it can be traced to its source and whether the system will tell someone when those conditions are no longer true.