Real-time production analytics

The challenge

Plant data is commonly rich at the control system level and thin everywhere else, reaching decision makers a day late as a summary in a spreadsheet. By the time a rate loss or quality excursion appears in the morning report, the shift that could have corrected it has gone home.

The obstacle is context rather than connectivity. Control system tags carry no business meaning, units differ between systems, and engineers maintain private spreadsheets with slightly different versions of the same calculation. Operations and engineering personnel require one live view of rate, availability and quality that everyone agrees on.

When it's time to act

The signs we commonly see when this initiative is due.

• Rate losses appear in the morning report, a full shift after they could have been corrected.
• Engineers keep private spreadsheets with slightly different versions of the same KPI.
• The control room can see everything and the daily meeting can see none of it.
• Units and tag names differ between systems, so no two people join the data the same way.
• Improvement targets argued about because the baseline moves with the calculation.

Our approach

• Understand where decisions are being made late and what that costs, with the operators, engineers and superintendents who live the delay.
• Assess the current data flows, definitions and integration maturity against your site standards and, where a common language is needed, published KPI references.
• Agree KPI formulas once and write them down, which ends the private-spreadsheet problem at the source. Existing site definitions usually come first, with references such as ISO 22400 available where a common language is needed.
• Integrate historian and control system data over OPC UA with read-only, secured data flow, respecting the boundary between operational and enterprise systems that models such as ISA-95 describe.
• Contextualise raw tags into equipment and product structures, normalise units, and validate streams against simple mass balance checks so bad sensors do not drive bad decisions.
• Build role-based dashboards: a live operations view for the control room, and shift and daily views for engineering and management.
• Add staged deviation alerts so a sustained rate loss pages the right person in minutes rather than appearing in tomorrow's report.

Tools and methods

The value it creates

• Operations sees rate, availability and quality while the shift is still running, so excursions get corrected instead of explained.
• The daily production meeting starts from one shared live dataset rather than three prepared spreadsheets.
• Consistent KPI definitions end reconciliation debates between operations, engineering and management, and give a defensible basis for improvement targets.

What changes

Where these initiatives fail

The failure modes we design against.

• Building dashboards before agreeing definitions, which projects the private-spreadsheet problem onto screens.
• Trusting sensors without validation, so a drifting instrument quietly drives decisions.
• Treating the control boundary casually: analytics must stay read-only and segmented.
• Leaving the daily routine unchanged, so the live view runs beside the old meeting instead of replacing it.

Common questions

Key terms

Plain-language definitions from our glossary for the concepts this page leans on.

Standards and further reading

Reference points we draw on where they suit the work. We also work to client internal standards and established site practice.

• OPC Unified Architecture, platform-independent industrial interoperability (OPC Foundation)
• ISA-95 enterprise-control system integration (ISA)
• ISO 22400-2:2014 KPIs for manufacturing operations management (ISO)
• ISO 14224:2016 Collection and exchange of reliability and maintenance data (ISO)
• Dimensional modelling techniques (Kimball Group)

Further reading

Articles and calculators on the methods behind this work.

• Digital transformation on a working mine site (article)
• OEE Calculator (tool)
• Time-Series Data Volume Calculator (tool)
• Oil and gas solutions (solution)
• Building software that runs a 24/7 operation (article)

Related projects