Alarms, trends & production reporting

Begin with the decisions people make

A useful report answers a specific operational question. Examples include how much was produced during a shift, why a line stopped or whether a process value remained within its operating range. Start with the intended reader and decision before choosing charts. An operator display, supervisor dashboard and management report serve different needs.

• Immediate equipment status for operators
• Shift output and interruption summaries for supervisors
• Historical trends for maintenance and process investigation
• Daily or weekly summaries for production planning

Define each measurement

Data is only comparable when its meaning is consistent. Decide where a production count comes from, how a running state is identified and what qualifies as downtime. Agree shift boundaries, timestamps, units and treatment of missing readings. A machine can be energised without producing, so status definitions should reflect the process rather than a convenient single bit.

• Production count source and reset behaviour
• Running, stopped, idle and fault-state definitions
• Scheduled versus unscheduled downtime
• Time zone, shift rollover and unit conventions

Make alarms actionable

An alarm should communicate an abnormal condition and help the operator understand what to check. Repeated alarms and vague messages can obscure the event that actually needs attention. Discuss message wording, priorities, acknowledgement and whether an event should be recorded without presenting as an alarm. The required behaviour depends on the process and existing platform.

• Equipment name and a clear description of the condition
• Relevant status or process values displayed nearby
• Alarm history and acknowledgement behaviour
• Review of recurring events and nuisance conditions

Connect the existing information sources

RS Automation’s published portfolio includes FactoryTalk Historian and transaction/reporting applications. A new reporting discussion should identify the installed versions, available tags and database or historian access. Direct machine values, archived process measurements and business records may require different collection methods. Refresh frequency and retention need to suit the operational question.

• Existing SCADA, historian and database platforms
• Tag lists, units and quality/status information
• Read-only access arrangements and report audience
• Update interval, retention period and export requirements

Validate before relying on the dashboard

Compare sample report results with independent plant records. Check shift changes, stopped periods, resets and missing data rather than validating only a normal running hour. Record the formulas and definitions used so that operators and managers interpret the same figure in the same way. Performance benefits should be measured after deployment rather than promised in advance.