If you run a production line or any manufacturing operation, you've probably heard the term OEE, or Overall Equipment Effectiveness. It's the single most widely used metric for understanding how well your production equipment is really performing.
But what exactly is OEE, how do you calculate it, and (most importantly) how do you use it to actually improve your output?
What Is OEE?
OEE is a percentage score that tells you how effectively your manufacturing equipment is being used. A score of 100% means you're producing only good parts, as fast as possible, with no downtime. In reality, no line runs at 100%, but tracking your OEE over time reveals where your biggest losses are hiding.
OEE was originally developed as part of Total Productive Maintenance (TPM) by Seiichi Nakajima in the 1980s. Today it's the standard metric used by food manufacturers, pharmaceutical companies, FMCG producers, and packaging operations worldwide.
The Three Components of OEE
OEE breaks down into three independent factors:
1. Availability
Availability measures the percentage of planned production time that the machine is actually running. It accounts for all events that stop production for an appreciable length of time, including breakdowns, changeovers, material shortages, and other unplanned stops.
Formula:
Availability = Run Time / Planned Production Time
If your shift is 8 hours (480 minutes) and you had 60 minutes of downtime (changeover, breakdown, waiting for materials), your Availability is:
420 / 480 = 87.5%
2. Performance
Performance measures whether the machine is running at its designed speed. Even when a machine is running, it might be operating slower than its rated capacity because of minor stops, slow cycles, or operator speed variations.
Formula:
Performance = (Total Units Produced / Run Time) / Rated Speed
If your VFFS machine is rated at 60 packs per minute, ran for 420 minutes, and produced 21,000 packs:
(21,000 / 420) / 60 = 50 / 60 = 83.3%
3. Quality
Quality measures the proportion of good units out of total units produced. Rejects, reworks, and defective products all reduce your Quality score.
Formula:
Quality = Good Units / Total Units Produced
If you produced 21,000 packs but 630 were rejected (seal failures, underweight, damaged):
20,370 / 21,000 = 97.0%
Want to crunch the numbers for your own line? Use our free OEE calculator to instantly calculate Availability, Performance, Quality, and OEE from your production data.
Calculating OEE
Once you have all three components, OEE is simply:
OEE = Availability x Performance x Quality
Using our example:
0.875 x 0.833 x 0.970 = 70.7%
What Does a Good OEE Score Look Like?
| OEE Score | Rating | What It Means |
|---|---|---|
| 85%+ | World Class | Top-tier performance, often cited as the benchmark |
| 60 to 85% | Good | Typical for well-managed operations, room for improvement |
| 40 to 60% | Average | Common for operations that don't actively track OEE |
| Below 40% | Low | Significant losses, usually a mix of all three factors |
Most packaging operations in New Zealand sit somewhere between 55% and 75% OEE. The important thing isn't hitting a magic number. It's understanding where your losses are and improving over time.
Step-by-Step OEE Calculation: A Complete Example
Let's walk through a full OEE calculation using realistic numbers from a VFFS (Vertical Form Fill Seal) packaging line running a single 8-hour shift.
The Scenario
A snack food manufacturer runs a VFFS machine rated at 80 packs per minute. The shift runs from 6:00 AM to 2:00 PM (480 minutes). Here's what happened during the shift:
- Planned breaks: 30 minutes (morning tea + lunch, excluded from planned production time)
- Changeover: 25 minutes (switching from 250g to 500g format)
- Breakdown: 18 minutes (jaw heater fault, maintenance callout)
- Material wait: 12 minutes (delayed film reel delivery from stores)
- Total units produced: 25,200 packs
- Rejected units: 380 packs (180 seal failures during startup, 200 underweight packs)
Step 1: Calculate Planned Production Time
Start by subtracting planned breaks from total shift time. Planned breaks are excluded because they are scheduled non-production time, not losses.
480 minutes - 30 minutes = 450 minutes planned production time
Step 2: Calculate Run Time (for Availability)
Subtract all downtime events from planned production time:
450 - 25 (changeover) - 18 (breakdown) - 12 (material wait) = 395 minutes run time
Availability = 395 / 450 = 87.8%
Step 3: Calculate Performance
The machine is rated at 80 packs per minute. During the 395 minutes of run time, it produced 25,200 packs:
Actual rate = 25,200 / 395 = 63.8 packs per minute
Performance = 63.8 / 80 = 79.7%
This tells us the machine was running about 20% below its rated speed, likely due to minor jams, slow cycles, or operator-induced slowdowns that didn't register as formal stops.
Step 4: Calculate Quality
Good units = 25,200 - 380 = 24,820
Quality = 24,820 / 25,200 = 98.5%
Step 5: Calculate OEE
OEE = 0.878 x 0.797 x 0.985 = 68.9%
Interpreting the Result
At 68.9%, this line is performing in the "Good" range but has clear room for improvement. The breakdown by component reveals where to focus:
- Availability (87.8%): The changeover and breakdown are the biggest losses. Reducing changeover time through SMED (Single-Minute Exchange of Die) techniques could recover 10 to 15 minutes per shift.
- Performance (79.7%): This is the weakest component. The machine is losing over 20% of its potential output to speed losses and minor stops, so it's worth investigating with detailed stop-reason tracking.
- Quality (98.5%): Already strong. The startup rejects are normal for a format change, and the underweight packs suggest a checkweigher calibration review.
Common OEE Mistakes Manufacturers Make
OEE is a straightforward metric, but it is surprisingly easy to calculate incorrectly or draw the wrong conclusions from the data.
1. Including Planned Downtime in Availability
Scheduled breaks, planned maintenance windows, and days when the line is not scheduled to run should be excluded from planned production time. If you include them, your Availability score will be artificially low, and you will be chasing "losses" that are actually deliberate choices.
2. Using the Wrong Rated Speed
Performance is calculated against the machine's rated (ideal) speed. Using a speed that is too conservative (such as the speed operators typically run at) masks real performance losses. Using a speed that is too aggressive (such as the maximum theoretical speed the machine was ever clocked at) makes performance look worse than it is. The right number is the manufacturer's stated sustained rate for the product format being run.
3. Ignoring Minor Stops
Stops under 2 to 3 minutes are often not recorded in manual OEE systems because operators don't bother writing them down. But minor stops (jams, sensor trips, brief upstream starvation) can account for 10 to 15% of total production time. This is one of the biggest reasons manual OEE overstates actual performance compared to automated tracking.
4. Averaging OEE Across Unlike Lines
Combining OEE from a high-speed cartoner running at 200 units per minute with a manual packing station into a single "site OEE" produces a meaningless number. OEE should be tracked per machine or per line, and compared only against the same line's historical performance.
5. Treating OEE as a Single Number Instead of Three
An OEE of 65% could mean three very different things: high availability but low performance, or strong performance with poor quality, or moderate scores across all three. Always look at the three components individually, because the aggregate number hides the actionable insight.
Practical Strategies to Improve Each OEE Component
Improving Availability
Availability losses come from downtime, both planned (changeovers) and unplanned (breakdowns). Practical strategies include:
- SMED changeovers. Apply Single-Minute Exchange of Die principles to reduce format changeover time. Separate internal tasks (machine must be stopped) from external tasks (can be done while running), and convert as many internal tasks to external as possible. Many packaging lines can cut changeover time by 30 to 50%.
- Preventive maintenance schedules. Track mean time between failures (MTBF) for common breakdown causes and schedule maintenance before failure occurs. Replacing a seal bar heater element during a planned stop is far less costly than an unplanned breakdown mid-run.
- Standardised startup procedures. Document and train operators on consistent startup sequences to reduce the time between "machine powered on" and "first good pack produced."
- Material staging. Ensure film reels, cartons, and other consumables are staged at the line before the shift starts. Waiting for materials is pure waste.
Improving Performance
Performance losses are the hardest to see without automated tracking, because the machine is technically "running," just not at full speed.
- Identify and address minor stops. Use automated stop tracking to capture every jam, sensor trip, and micro-stop. Pareto-chart the top causes and address them systematically. A VFFS machine that jams 30 times per shift for 10 seconds each loses 5 minutes. That's invisible on a paper log, but clearly visible in PLC data.
- Review speed settings. Are operators running below rated speed "because it runs better slower"? Sometimes this is valid (poor-quality film, difficult product), but often it's a habit that can be challenged with data.
- Upstream and downstream constraints. If the machine starves because the upstream feeder cannot keep up, or backs up because the downstream case packer is too slow, the line's overall performance suffers. OEE data helps identify which machine is the bottleneck.
Improving Quality
Quality losses are usually the smallest OEE component for packaging lines, but every rejected pack is wasted material and lost throughput.
- Startup reject reduction. Track the number of rejects produced during each startup and changeover. If the machine consistently produces 50 bad packs before settling, investigate whether pre-heating, film tension adjustment, or recipe parameters can reduce that number.
- Checkweigher and metal detector calibration. Ensure detection equipment is calibrated correctly. Too sensitive and you reject good product; too loose and you miss defects.
- Film and material quality. Poor-quality packaging film causes seal failures and jams. Track reject rates by film batch to identify supplier quality issues.
OEE Benchmarks by Industry
OEE varies significantly by industry due to differences in product complexity, changeover frequency, batch sizes, and regulatory requirements.
| Industry | Typical OEE Range | Notes |
|---|---|---|
| Automotive | 80 to 90% | Long runs, high automation, mature TPM programmes |
| FMCG / Consumer Goods | 65 to 80% | High-speed lines, moderate changeover frequency |
| Food Manufacturing | 55 to 65% | Frequent changeovers, perishable materials, strict hygiene requirements |
| Beverage | 60 to 75% | High-speed filling, CIP (clean-in-place) downtime |
| Pharmaceutical | 45 to 55% | Small batches, extensive cleaning validation, regulatory holds |
| Bakery | 50 to 60% | Product variability, dough handling challenges |
| Meat Processing | 45 to 55% | Variable raw material, manual intervention, sanitation requirements |
These ranges are guides, not targets. A food manufacturer consistently achieving 65% OEE with reliable data is performing well. A manufacturer claiming 90% OEE on a manual tracking system probably has a data accuracy problem.
For New Zealand food manufacturers specifically, the 55 to 65% range is realistic and typical. Operations that introduce automated OEE tracking often discover their actual OEE is 5 to 15% lower than their manual records suggested. That can be uncomfortable at first, but it's valuable because it reveals the real improvement opportunities.
The Six Big Losses
OEE is designed to surface what's known as the Six Big Losses, the most common causes of manufacturing inefficiency:
Availability Losses:
- Equipment Breakdowns: unplanned stops such as mechanical failure and PLC faults
- Setup & Changeovers: time spent switching between products or formats
Performance Losses: 3. Minor Stops: brief interruptions such as jams, sensor trips, and upstream starving 4. Reduced Speed: running below rated capacity (often hidden and underestimated)
Quality Losses: 5. Start-up Rejects: defective products during warm-up or changeover 6. Production Rejects: defects during normal production, such as seal failures and weight errors
Manual OEE vs Automated OEE
Traditionally, OEE is calculated from paper-based shift logs. An operator writes down when the machine stopped, why it stopped, how many units were produced, and how many were rejected. A supervisor then enters this into a spreadsheet.
The problem? Manual OEE data is slow, inaccurate, and inconsistent. Operators often round numbers, miss short stops entirely, and classify downtime reasons inconsistently. Studies show manual OEE can be 10 to 20% different from actual OEE.
Automated OEE solves this by reading data directly from the machine's PLC (Programmable Logic Controller). The PLC already knows the machine state, production counts, and reject counts. It just needs a way to get that data to a dashboard.
This is exactly what solutions like LineConnect+ do. By subscribing to OPC-UA data from your production line PLCs, production data flows automatically into cloud dashboards, calculating OEE every hour without any manual data entry. For a deeper look at how this compares to traditional approaches, see our guide to SCADA vs cloud monitoring.
How to Improve Your OEE
Once you're tracking OEE automatically, you can start making targeted improvements:
Focus on your biggest loss first. Is it Availability (too much downtime), Performance (running too slow), or Quality (too many rejects)? The OEE breakdown tells you exactly where to look.
Track downtime reasons. Don't just know that the machine stopped. Know why it stopped. Categorise stops by reason code and tackle the top offenders.
Monitor trends, not snapshots. A single hour of OEE data is not useful on its own. Look at daily, weekly, and monthly trends. Is your changeover time improving? Are breakdowns becoming less frequent?
Set realistic targets. If you're at 55% OEE, don't aim for 85% overnight. Target 60%, then 65%, and build momentum.
Make the data visible. Put OEE dashboards where operators and supervisors can see them in real time. Visibility drives accountability and faster response to issues.
OEE for New Zealand Manufacturers
For NZ food manufacturers and packaging operations, OEE is particularly valuable because:
- Margins are tight. Even a 5% improvement in OEE can translate to significant revenue gains on high-volume lines.
- Labour is expensive. Understanding whether downtime is equipment-related or process-related helps you allocate resources effectively.
- MPI compliance matters. Linking OEE data with batch traceability gives you a complete picture of what was produced, when, and at what quality level.
Getting Started with Automated OEE
If you're currently tracking OEE manually (or not at all), the path to automated OEE monitoring is simpler than you might think.
- Check whether your packaging machines support OPC-UA or have a PackML-compatible PLC interface.
- Identify the key data points: machine state, good count, reject count, and rated speed.
- Connect a lightweight edge gateway to your PLC network.
- Start receiving real-time OEE data in cloud dashboards within days, not months.
Frostbyte Pro's LineConnect+ module is designed specifically for this. It connects to your production line PLCs via OPC-UA, calculates OEE automatically, and presents it alongside your inventory and production data in a single platform.
Ready to move from manual OEE tracking to automated, real-time dashboards? Try our free OEE calculator to see where your line stands, then upgrade to automated tracking with Frostbyte Pro and LineConnect+. Connect to your PLCs via OPC-UA and see Availability, Performance, and Quality calculated automatically. Start your free trial today, no credit card required.
Want to see what automated OEE monitoring looks like for your packaging line? Get in touch. We'd love to chat about your line configuration.