Quick Answer: Test reports and certifications show that a battery pack design can meet its specification. They say far less about whether a supplier can build that pack reliably, week after week, at the volume you need.
Five production line metrics fill that gap:
-
Overall equipment effectiveness (OEE) with its full breakdown.
-
Unplanned downtime and its causes.
-
Cycle time against the ideal rate.
-
First-pass quality at each station and changeover time.
-
Trends over several months, plus an explanation of how each figure is collected; not a single headline number.
Key Takeaways
-
Line performance is a leading indicator of delivery and quality risk, so it belongs in supplier qualification alongside test evidence.
-
OEE is only useful when it is split into availability, performance and quality, because two lines with the same score can carry very different risks.
-
Downtime causes, cycle time and station-level quality show where a supplier is losing capacity before it affects your orders.
-
Changeover time matters most for custom, low-volume packs that share a line with other products.
-
Data collection methods are as important as the numbers themselves; manual records tend to understate losses.
Part 1: Why Line Performance Belongs in Supplier Qualification
Most OEM qualification work focuses on design evidence. Buyers review cell and pack test reports, check certifications and confirm that the configuration is locked before release. While this work is essential, it only answers one question: can this pack meet the specification? What it fails to touch on is whether the supplier can keep building packs at the agreed rate and quality.
The gap between the two can be expensive. In June 2024, BMW withdrew a cell order worth around €2 billion from Northvolt. The manufacturer reportedly cited production delays and poor cell quality as key reasons behind the decision. Later that year, Reuters reported that the company had been consistently missing weekly production goals, delivering 22,000 shippable cells in one October week against a target of 30,000. While Northvolt was a cell maker operating at gigafactory scale, not a small capacity pack assembler, the takeaway remains the same; the difficulties faced surfaced on the production line, they weren’t a result of bad design.
Unplanned stoppages are costly across the entirety of the manufacturing industry. According to Siemens’ True Cost of Downtime 2024 report, the world’s 500 largest companies lose 11% of their revenues to unplanned downtime, while the average plant loses 27 hours of productivity per month.
Production line metrics give OEMs an early view of these risks. Yield and process capability are usually reviewed only once during the design freeze and validation stage. By analyzing the five metrics below, operators can continuously investigate how the line performs daily, in real world conditions, to gain a better understanding where risks and faults might occur.
Part 2: The Five Metrics
2.1 OEE, and Why to Ask for the Breakdown
OEE measures how much of a line’s planned production time actually results in good output by highlighting three factors:
-
Availability: The share of planned time the line is running rather than stopped.
-
Performance: How close the running speed is to the ideal rate.
-
Quality: The share of units that pass first time without rework or scrap.
These three factors are multiplied together to produce an OEE score displayed as a percentage, with a 100% score representing a flawless production process.
Seiichi Nakajima, who introduced the measure, described a score above 85% as world class, while most factories average around 40% to 60%. A headline figure on its own tells a buyer very little, but when OEE is viewed as a diagnostic tool, not an arbitrary target, the figure can help to inform practical improvements.
The table below shows why. Consider two hypothetical pack lines that report the same score:
|
Line |
Availability |
Performance |
Quality |
OEE |
Main risk for the OEM |
|---|---|---|---|---|---|
|
Line A |
80% |
95% |
95% |
72% |
Frequent stoppages that can delay deliveries |
|
Line B |
96% |
94% |
80% |
72% |
High rework and scrap that put consistency at risk |
Both lines score 72%, yet a buyer worried about lead times should look harder at Line A, while one worried about field failures should focus on Line B.
For OEMs, ask that OEE monitoring be performed on the specific line that will build your pack, split into all three factors across the last six to twelve months. It is also worth asking what counts as planned production time, since excluding breaks, maintenance or changeovers can make a score look higher than the line’s real output.
2.2 Unplanned Downtime and Its Top Causes
Availability losses come mostly from unplanned stops. On a pack line, these often cluster around a few stations, namely cell tab welding, BMS programming and end-of-line testing where a single faulty fixture or tester can hold up everything upstream.
Rather than asking for a total, ask for downtime hours per month together with the top five causes ranked by lost time. A supplier who can name its biggest sources of downtime and show what it has done about the largest one, understands their own process; a supplier who can only quote a single monthly figure may not yet be tracking causes closely enough to fix them.
2.3 Cycle Time Against the Ideal Rate
Performance losses are less visible than stoppages. A line can be running all shifts and still underdeliver because stations run below their ideal speed or pause repeatedly for a few seconds at a time. These short stops rarely appear in downtime logs, but they all add up.
A simple example shows the scale. If a line’s ideal cycle time is 60 seconds per pack but the actual average is 75 seconds, performance sits at 80%. Over a five-shift week, that means roughly one full shift is lost to slow running alone.
Considering this, OEMs should ask for the ideal and actual cycle time at the bottleneck station and how the line’s capacity compares with your forecast demand. A supplier running close to 100% of its practical capacity has little room to absorb a demand spike or recover from even a single bad week.
2.4 First-Pass Quality at Each Station
Quality losses in battery production can be substantial. An Accenture and Fraunhofer FFB whitepaper on cell manufacturing notes that scrap rates can vary between 2% and 25%. Pack assembly involves different processes, but the principle holds; losses that never reach the customer still affect cost, capacity and consistency.
The key distinction is between end-of-line pass rate and first-pass quality. End-of-line testing confirms that each shipped pack meets its release criteria. A high pass rate there can hide a line that reworks a large share of units before they reach the tester. OEMs must ask for first-pass rates at each major station, such as welding, BMS programming, insulation and final test, along with the most common defect types. Station-level data shows where variation enters the process which is far more useful than a single overall percentage.
2.5 Changeover Time for Low-Volume, Custom Builds
Custom battery packs are often built in smaller batches on lines shared with other products. Each switch between products can mean new fixtures, weld programs, tester profiles and BMS firmware, with every minute spent changing over being a minute the line is not producing.
In the real world, OEE can vary significantly depending on product mix and process complexity; variables that differ wildly on high-mix pack lines. OEMs should ask their suppliers how long a typical changeover takes for a product similar to the one they’re considering, how often the line switches products and what the minimum practical batch size is. It’s also worth confirming whether changeovers are counted as planned or unplanned time, as that choice can influence the OEE figure considerably.
Part 3: How the Data Is Collected Matters
Every metric above depends on the quality of the underlying records. Two suppliers can report similar numbers while measuring them in very different ways.
One of the most useful questions is whether figures come from manual shift logs or from automated OEE monitoring systems connected to machines, since handwritten records tend to miss short stops and round off downtime. There are pros and cons to each approach, but buyers should understand which one is behind the numbers they are shown.
Other questions worth asking include:
-
Are definitions, such as planned time and ideal cycle time, consistent across lines and sites?
-
Can the supplier share trend data over several months, not just a recent snapshot?
-
Who reviews the figures internally, and how often do they lead to corrective actions?
-
Can line data be linked to serial numbers if a quality issue needs to be traced?
Part 4: A Supplier Question Checklist
The table below can be used during supplier audits or quarterly business reviews.
|
Metric |
What to Ask |
A Good Answer |
A Red Flag |
|---|---|---|---|
|
OEE |
OEE for our line, split into three factors, over 6-12 months |
A breakdown with a clear trend and defined planned time |
A single figure with no breakdown or history |
|
Unplanned downtime |
Hours per month and the top five causes |
Ranked causes with actions taken on the largest |
A monthly total with no cause data |
|
Cycle time |
Ideal vs actual cycle time at the bottleneck |
Known bottleneck with spare capacity against our forecast |
No ideal rate defined or no spare capacity |
|
First-pass quality |
First-pass rates by station and main defect types |
Station-level data with defect trends |
Only an end-of-line pass rate |
|
Changeover time |
Typical changeover for a product like ours |
Measured times and a plan to reduce them |
Changeovers left out of OEE without explanation |
|
Data collection |
Manual logs or machine data? |
Clear method and consistent definitions |
Figures compiled by hand at month end |
FAQs
Is 85% OEE a Fair Target for a Battery Pack Line?
It’s a useful benchmark rather than a pass mark. High-mix lines building custom packs face more changeovers than dedicated high-volume lines, so a lower score is not automatically a concern. Steady improvement over time usually says more than the absolute number.
Should OEMs Compare OEE Between Suppliers?
Only with care. Suppliers often define planned time and ideal rates differently, making direct comparisons unreliable. Ultimately, the most important benchmark for any manufacturer is its own historical performance.
What If a Supplier Won’t Share Production Data?
Some suppliers treat line data as commercially sensitive. Asking for summarized trends or reviewing figures on site during an audit can be a workable middle ground. A complete refusal to discuss line performance is worth noting during qualification.
How Often Should OEMs Review These Metrics?
Reviewing them at qualification and then quarterly works for many programs. More frequent reviews make sense during ramp-up, after a design change or when delivery performance slips.
Do These Metrics Replace Test Reports and Audits?
No. Test reports, certifications and process audits confirm that a pack is safe and meets its specification. Production line metrics add a different view; whether the supplier can keep delivering that pack consistently once volume production begins.
