Glossary · IIoT, data and AI
Analytics (industrial)
Also known as: Industrial analytics, Operational analytics
German: Analytik
In industrial automation, analytics is the systematic use of machine, process and business data to describe what happened, diagnose why, predict what is likely to happen and recommend actions. It spans simple KPI dashboards as well as statistical and machine learning models.
- IIoT
In one sentence
Industrial analytics turns machine, process and business data into descriptions, diagnoses, predictions and recommended actions.
Example
A packaging line's analytics dashboard combines PLC counters and downtime reasons to show that most stops occur at the labeler during format changes.
How it applies
- Engineering: Analytics needs data that is time-stamped, contextualized and of known quality. Plan Data pipelines, tag naming and Metadata early, not after the plant is running.
- Operation: Descriptive and diagnostic analytics support shift reports and root-cause analysis; predictive and prescriptive analytics support maintenance and scheduling decisions.
- Documentation: Document each KPI and model: data sources, calculation rules, units, update rate and known limitations. Operators need to know what a value means and when not to trust it.
Analytics vs. data analytics
Analytics is used here as the umbrella for the capability in a plant or organization, including the tools, roles and decisions it feeds. Data analytics is the concrete process of cleaning, transforming and analyzing data sets. In everyday use the two terms are often interchangeable.
Analytics results inform decisions; they do not replace the verification of safety functions or the approval of process changes through Change control.