OEE dashboard in Power BI: measures and checks

An OEE dashboard in Power BI should explain which production losses sit behind overall equipment effectiveness: availability, performance and quality. Show the components next to lost minutes and good output, using one consistent schedule boundary. The arithmetic should still reconcile when the visual layout changes.

Microsoft distinguishes a one-page Power BI dashboard from the detailed reports behind it. This example specifies the business measures and checks for that summary. Covirage is not affiliated with Microsoft or Power BI.

Establish the production boundary

Take a closed operating export with a run identifier, date, line, product, planned production minutes, stop minutes, total units and first-pass good units. Add the approved ideal cycle time in minutes per unit. A separate schedule export establishes when production was intended to run.

Exclude periods when production was not scheduled before determining planned production time. Keep stops occurring within that production window, including scheduled changeovers, in stop time. The agreed boundary must remain consistent across lines and periods; removing an inconvenient stop from the denominator changes the question.

Vorne's OEE calculation guide defines the three component formulas and the equivalent direct calculation:

Availability = Run minutes / Planned production minutes

Performance = Ideal cycle minutes × Total units / Run minutes

Quality = First-pass good units / Total units

OEE = Availability × Performance × Quality

Direct OEE check = Ideal cycle minutes × First-pass good units / Planned production minutes

Run minutes equal planned minutes less stop minutes. Ideal cycle time is a technical production standard, not a budget allowance inflated to make the score easier to achieve. The general OEE dashboard guide covers the wider operating review; the focus here is a set of measures a business owner can verify before presenting them.

A worked example: two Ohio production lines

An illustrative Ohio packaging manufacturer compares two lines on September 15, 2026. Both make the same container, using an approved ideal cycle time of 0.5 minute per unit. Each has 450 scheduled production minutes after excluded nonproduction breaks.

Calculate Line A's availability as 400 / 450 = 88.89%, performance as (0.5 × 700) / 400 = 87.50%, and quality as 665 / 700 = 95.00%. Multiplying the unrounded ratios gives 73.89% OEE. The direct check is (0.5 × 665) / 450 = 73.89%.

The combined performance is (0.5 × 1,300) / 760 = 85.53%. Combined OEE is (0.5 × 1,235) / 900 = 68.61%. Percentages are displayed to two decimals; calculations retain their original precision.

These lines share the same ideal cycle time. With different products, calculate fully productive minutes for each run using its own ideal time, sum those minutes, and divide by total planned minutes. Do not apply an arbitrary common cycle time to unlike products.

Prove the total with a loss bridge

A score gives the production manager a summary. A bridge from planned time to fully productive time gives the manager a work list.

Speed loss minutes = Run minutes − Ideal cycle minutes × Total units

Quality loss minutes = Ideal cycle minutes × (Total units − First-pass good units)

The proof is 140 + 110 + 32.5 + 617.5 = 900 minutes. No minute belongs in two loss categories. Line B has the larger stop loss, which is a reason to examine its stop register before discussing equipment replacement or staffing.

For a dollar comparison, suppose finance applies a standard conversion value of $2 per ideal production minute. Quality losses represent 32.5 × 2 = $65 of conversion value; all measured losses represent (900 − 617.5) × 2 = $565. These are time valuations under the stated assumption. They are not invoices, avoidable expenses or promised savings.

Change one input: more first-pass good output

Keep Line B's schedule, stop minutes, total production and ideal cycle time unchanged. Change only its first-pass good count from 570 to 585, reflecting fewer initial rejects.

Availability and performance stay unchanged. OEE improves by 0.83 percentage point, calculated before rounding, and the quality loss valuation falls by $15. The bridge still closes: 140 + 110 + 25 + 625 = 900 minutes. A quality improvement should not be presented as a reduction in downtime when the stop register has not changed.

Keep the review tied to accountable actions

Put the component scores and the loss bridge beside a line comparison. Production owns counts and schedule boundaries, maintenance owns stop reason investigation, and quality owns the first-pass acceptance definition. Finance owns any dollar valuation applied to the lost time.

Use production dashboards for the wider output view and a Power BI operations dashboard for decisions spanning several functions. If energy is also under review, the Power BI energy dashboard keeps utility consumption and tariffs separate from equipment effectiveness.

Where it goes wrong

• Averaging the displayed percentages. Calculate totals from productive and planned minutes. Rounded component percentages do not preserve the original calculation exactly.
• Repaired units become first-pass good. Preserve the initial inspection outcome even when rejected parts later pass after rework.
• Cycle standards drift. Keep the product-specific approved standard and its effective date; a standard change can move OEE without an operating change.
• Counts and time cover different windows. A run finishing after the reporting cutoff needs an explicit treatment, rather than a count in one period and its time in another.
• A higher score becomes a savings claim. Verify whether recovered time changes actual spend or usable output before attaching a financial benefit.

Check your production export in Covirage

Bring run-level production counts, the schedule, stop records and approved ideal cycle times. Covirage maps the columns, and its tools compute and check the component measures and totals; the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own data.

See manufacturing analytics for the supporting business context.