Measuring the Plant in Power BI: OEE, Scrap and Cost Variance Without a Separate System

Consider the following scenario. Your production manager wants to know why yesterday's output missed the target. Finance wants to understand why manufacturing costs are increasing. Quality wants to know which products are generating the most scrap.

Business Central already holds some of the information needed to answer those questions. The problem arises when production, quality and finance teams are looking at different reports. Each department sees part of the picture, but nobody has a consistent view of plant performance.

Before investing in a manufacturing execution system (MES), examine what Business Central captures and what Power BI can do with it.

A useful manufacturing Power BI OEE and scrap dashboard does not require every machine to be connected. But it does require accurate production data, consistent definitions and an understanding of what Business Central cannot measure.

What Business Central already captures about production

Business Central Premium includes manufacturing functionality for production orders, bills of material, routings, work centres, machine centres and capacity planning.

Depending on your configuration and posting practices, that information can include:

• Finished quantities and scrap quantities.
• Material consumption.
• Setup time and run time.
• Production order costs.
• Work centre and machine centre activity.
• Expected and actual production costs.

You can compare planned quantities against output, examine consumption variances and investigate production orders that cost more than expected. Business Central also maintains capacity ledger entries and item ledger entries that can support manufacturing analysis.

Note: Business Central can only report accurately on production activity your employees actually record.

• If operators enter one total output quantity at the end of a shift, you will not have reliable hourly production trends.
• If scrap quantities are recorded without meaningful reasons, you can calculate scrap rates but cannot explain why the scrap occurred.
• If actual machine time is never posted, Power BI cannot reconstruct it from a routing estimate.

Start by examining what is captured today, not what the system is theoretically capable of storing.

OEE, and why most companies calculate it wrong

Overall Equipment Effectiveness, or OEE, measures how effectively manufacturing equipment performs during planned production time. The standard calculation combines three factors:

OEE = Availability × Performance × Quality

1. Availability measures actual operating time against planned production time.
2. Performance compares actual output against the theoretical output possible during operating time, using an established ideal cycle time.
3. Quality measures good output against total output.

Consider a production shift where the planned production time is 480 minutes. The machine runs for 420 minutes. Its ideal cycle time is one minute per unit. It produces 380 units, of which 360 are good.

• Availability: 420 ÷ 480 = 87.5%
• Performance: 380 ÷ 420 = 90.5%
• Quality: 360 ÷ 380 = 94.7%
• OEE: approximately 75%

Here are the problems companies frequently run into:

• They calculate OEE using scheduled shift hours without accounting for planned production time correctly.
• They use estimated routing times as actual machine runtime.
• They count all production output as good output because scrap is recorded later.
• They combine several machines into one OEE figure without accounting for different cycle times or operating conditions.

Each of those mistakes changes the result. Business Central can hold posted production quantities, scrap and time information. It does not automatically provide a complete, independently measured record of every machine stop, ideal cycle time and operating condition. That distinction determines whether you can calculate genuine OEE or only an approximation.

Do not label a production-efficiency estimate as OEE unless the underlying data supports the calculation.

Scrap and yield reporting from production journals

Scrap reporting is often the easiest place to begin because Business Central already supports recording scrap during production. Operators can record output quantities and scrap quantities through production journals, subject to the production process and configuration. Business Central also supports scrap codes that can identify reasons for scrap. That gives you a foundation for two useful measures:

• Scrap rate = Scrap quantity ÷ Total produced quantity
• Yield = Good output ÷ Total produced quantity

Consider a production operation that produces 1,000 units, of which 45 are scrapped. The scrap rate is 4.5% and the yield is 95.5%.

Those calculations assume good and scrap quantities are captured consistently at the same production stage. In multistage manufacturing, a component may be scrapped during one operation while another unit is rejected during final inspection.

You need to decide whether the dashboard measures scrap by operation, production order, item or finished output. Otherwise, quantities can be double-counted or compared incorrectly. A useful Power BI report should show scrap by item, work centre, production order and reason code. Then add cost.

Note: Scrapping ten inexpensive components is not necessarily as significant as scrapping one high-value assembly.

Business Central's costing information can help quantify the financial impact, provided the cost allocation and posting processes are appropriate. The first improvement is often not a better dashboard. It is getting operators to record scrap accurately and consistently.

Standard vs. actual cost variance, and reading it correctly

Manufacturing cost variance deserves its own reporting model. Business Central supports production costing and standard costing, including calculations based on materials, capacity and manufacturing overhead. It also records actual consumption and output costs.

But the interpretation depends on the item's costing method, production-order status and whether cost adjustment processes have been completed. For standard-costed production, you can examine differences between expected or standard costs and actual production costs.

Consider a product with a standard manufacturing cost of $120. The completed production order incurs $132 per finished unit. The difference is $12 per unit, or 10% above standard.

That tells you the order cost more than expected. It does not tell you why. Possible causes include additional material consumption, higher input costs, excess setup time, inefficient production or overhead differences. Power BI can help separate those patterns when the underlying data supports the analysis.

But do not compare unfinished production orders against completed orders without accounting for work in progress. Business Central's costing and adjustment processes also matter. Costs may change after initial posting as inventory costs are adjusted. A dashboard refreshed before those adjustments are complete may disagree with finance.

The purpose of cost variance reporting is to identify what changed, not simply to display a red percentage.

Where you genuinely need shop-floor data Business Central does not hold

There is a limit to what posted ERP transactions can tell you. If you need exact machine downtime by minute, automatic cycle counts, machine-state monitoring or sensor-based performance information, Business Central is not a substitute for a shop-floor data collection system.

The same applies to detailed production traceability where every event, inspection and machine condition must be captured automatically. You may need an MES, machine connectivity, specialized quality software or another operational data source. Power BI cannot create missing observations. It can analyze them once they exist.

The distinction becomes especially important for OEE. A production journal might record 420 minutes of runtime. It may not explain whether the remaining 60 minutes were caused by equipment failure, material shortages, changeovers or operator availability. Without downtime reasons, management knows production time was lost but cannot reliably identify the cause.

Building something useful without an MES

You do not need to begin with automated machine connectivity. Start with one production line or work centre. Confirm that operators record output, scrap, setup time and run time consistently. Use a manageable reporting period, such as the previous four completed weeks. Then build a Power BI model using the relevant Business Central production and costing data.

Establish relationships between production orders, items, operations, work centres and dates.

• For scrap reporting, agree on the denominator and treatment of rework.
• For cost variance, agree on when production costs are considered final.
• For OEE, document which inputs are measured and which are estimates.

If actual machine downtime is unavailable, report production efficiency and scrap separately rather than presenting an unreliable OEE figure.

The three charts a plant manager actually opens

Most plant managers do not need a dashboard with 25 visualizations. They need to know where performance changed and what requires attention. These three views connect production performance, quality and financial results.

1. Production performance by line or work centre. Show planned versus actual output, good units, scrap and available time. Include OEE only where availability, performance and quality inputs are sufficiently reliable. Let the manager identify which line missed its target and investigate the underlying production orders.
2. Scrap rate and scrap cost by reason. Show scrap quantity, scrap percentage and estimated financial impact. Allow filtering by item, production order, operation and work centre. A trend over several weeks is more useful than a single total. The goal is to identify recurring causes that deserve corrective action.
3. Manufacturing cost variance by production order. Show standard or expected cost against actual cost for completed orders. Break out material, capacity and overhead differences where the available data and costing configuration support it.

Highlight the orders with the largest financial impact, not merely the highest percentage variance. A 30% variance on a small order may matter less than a 4% variance on a high-volume product.

Where to start

• Before buying an MES or commissioning a large manufacturing analytics project, audit what your plant already records in Business Central.
• Choose one production line.
• Review output postings, scrap quantities, time entries, routing standards and production costs.
• Identify what is reliable, what is estimated and what is missing.
• Then build the first dashboard around those facts.

Qixas Group works across Dynamics 365 Business Central, Power BI, Microsoft Fabric and Dynamics 365 Supply Chain Management. We help manufacturers improve the systems they already have and identify when additional technology is genuinely necessary.