Machine Capability Study — definition
Machine capability study is a statistical assessment of a machine's ability to consistently produce output within specified tolerances under short-term controlled conditions.
The study typically involves running a defined number of consecutive parts or cycles, measuring key characteristics, and calculating capability indices such as Cm or Cmk. It isolates machine-inherent variation from broader process variation captured by longer-term process capability studies.
Why it matters to industrial buyers
A documented capability study gives buyers objective evidence that a machine meets quality requirements before accepting it into production.
Key reference points
Typical sample size
Short-term capability studies commonly use 50 or more consecutive parts or cycles produced without interruption, adjustment or operator change.
Common threshold
A Cmk value of 1.33 or higher is commonly used as an acceptance threshold; 1.67 is frequently required for safety-critical or regulated characteristics, though requirements vary by industry.
Index pair
Cm describes the spread of the machine relative to the tolerance width; Cmk additionally penalises a mean that sits off the tolerance centre. Contracts should name which index is being accepted.
Measurement system
Capability figures are only as good as the gauge. A measurement system analysis is commonly run first, since gauge variation consuming a large share of tolerance can make a capable machine look incapable.
Commonly confused with
process capability study (Cpk)
A machine capability study isolates equipment variation over a short uninterrupted run; process capability includes longer-term variation from materials, operators, tooling and environment, and is measured over days or weeks.
factory acceptance test
A FAT confirms the machine functions and meets the specification at the supplier's works; a capability study quantifies statistically how tightly it holds a measured characteristic.
performance / reliability run
A performance run measures throughput and uptime over an extended period; a capability study measures dimensional or process variation, not availability.
How it is used in practice
A machine capability study on a filling line confirmed a Cmk of 1.5 for fill volume across 50 consecutive containers before the line was released for production.
This guide covers how a machine capability study is actually run, how Cm and Cmk are calculated with a worked example, what thresholds to accept, when the study applies and when it does not, and the exact wording to place in an equipment RFQ so capability is a supplier obligation rather than a post-delivery argument.
How the study is run
A capability study is deliberately narrow: one machine, one characteristic set, one uninterrupted run, no adjustments. The point is to isolate the machine's own variation from everything else that moves a process.
- Fix the conditions: one operator, one material batch, one tool set, one program, stable utilities and ambient conditions.
- Run at least 50 consecutive parts or cycles with no adjustment, no tool change and no intervention.
- Measure each part for the named critical characteristics, in production order, with a gauge that has passed a measurement system analysis.
- Check the data before the indices: plot the values in run order to confirm there is no trend, step or drift, and confirm the distribution is roughly normal.
- Calculate Cm and Cmk per characteristic and record the sample size, gauge, conditions and raw data in the study report.
A capability number reported without the raw data, the sample size and the measurement system is not evidence. Ask for the dataset, not the summary.
Worked example: fill volume on a liquid filler
Specification: 500 ml nominal with a tolerance of ±5 ml, so the tolerance limits are 495 ml and 505 ml and the tolerance width is 10 ml. Fifty consecutive containers are filled and weighed.
| Item | Value | How it is derived |
|---|---|---|
| Tolerance width | 10 ml | 505 ml upper limit minus 495 ml lower limit |
| Sample mean | 501.0 ml | Average of the 50 measured fills |
| Sample standard deviation | 1.0 ml | Standard deviation of the same 50 values |
| Cm | 1.67 | Tolerance width ÷ (6 × standard deviation) = 10 ÷ 6.0 |
| Distance to nearer limit | 4.0 ml | 505 ml upper limit minus the 501.0 ml mean |
| Cmk | 1.33 | 4.0 ÷ (3 × 1.0) |
| Cmk if re-centred at 500 ml | 1.67 | 5.0 ÷ (3 × 1.0) — same machine, mean moved to centre |
The machine is precise enough for 1.67, but the one-millilitre offset in the mean costs a full capability grade. That distinction — spread versus centring — decides whether the fix is an adjustment or a machine problem.
Reading the result
Where Cm is comfortably above target but Cmk is not, the machine is capable and mis-centred. Where both are low, the variation is real and no adjustment will fix it.
| Cmk | Interpretation | Typical action |
|---|---|---|
| Below 1.00 | Machine spread exceeds the tolerance; scrap is expected in normal running | Reject; investigate machine, fixture and gauge before re-test |
| 1.00 – 1.32 | Marginal; small shifts in condition will produce out-of-tolerance parts | Conditional acceptance at most, with a corrective plan and re-test |
| 1.33 – 1.66 | Commonly accepted for general industrial characteristics | Accept where the RFQ set 1.33; monitor with SPC in production |
| 1.67 and above | Commonly required for safety-critical or regulated characteristics | Accept; retain the dataset as the baseline for later process capability |
When it applies — and when it does not
- Applies to measurable, continuous characteristics: dimensions, fill weight or volume, torque, temperature, seal strength, coating thickness.
- Applies to new or relocated machines, after major retrofit, and after a change of tooling, control system or material grade.
- Does not apply cleanly to attribute results — pass/fail, present/absent — which need attribute agreement studies or much larger samples instead.
- Does not substitute for a performance run: a machine can be highly capable dimensionally and still fail on availability, changeover time or throughput.
- Of limited value on processes with a deliberate trend, such as tool wear, unless the study is designed and analysed for that drift.
What to write into the RFQ
Capability is cheap to specify before contract and expensive to argue about after delivery. The enquiry should name the criteria so that every bidder prices the same test.
- The critical characteristics to be studied, with the drawing or specification reference and the tolerance for each.
- The required index and level per characteristic — for example Cmk ≥ 1.33, or ≥ 1.67 where safety-related.
- Sample size and run conditions: consecutive parts, no adjustment, buyer-supplied or supplier-supplied material, which shift and which program.
- Where the study is performed: at FAT, at SAT, or both — and on whose material.
- The measurement system: the gauge to be used and whether an MSA or gauge R&R result must be submitted with the study.
- Consequences of failure: who pays for re-work and re-test, how many re-tests are allowed, and whether capability is a condition of the acceptance milestone payment.
- Deliverables: raw measurement data, the capability report, the run chart and the conditions log — not just the index values.
If capability is tied to a payment milestone, state it in the payment schedule as well as the technical annex, or the clause is difficult to enforce.
Frequently asked questions
What is a machine capability study?
A short-run statistical check of how consistently a machine holds a specified characteristic. A defined number of consecutive parts is produced under fixed conditions, each is measured, and the spread and centring of the results are expressed as Cm and Cmk against the tolerance.
How is Cmk calculated?
Cm is the tolerance width divided by six standard deviations of the sample. Cmk takes the smaller of the distances from the mean to the upper and lower tolerance limits, divided by three standard deviations, so a machine that is precise but off-centre scores lower on Cmk than on Cm.
What Cmk value should be accepted?
1.33 is the most commonly cited acceptance level for general industrial characteristics, with 1.67 frequently required for safety, regulatory or high-consequence characteristics. The value should be stated per characteristic in the contract, not assumed.
How many parts are needed?
50 consecutive parts is the most widely used minimum for a short-term study; some standards and customer specifications call for 100 or more. Fewer parts widen the confidence interval, so a marginal result on a small sample is not strong evidence of capability.
When is a capability study performed?
Typically at factory acceptance test on the supplier's floor and again at site acceptance on the buyer's production material, before formal production release. Repeating it on site matters because material, utilities and ambient conditions differ.
Does a capability study replace ongoing quality control?
No. It confirms initial capability under controlled conditions; statistical process control monitors sustained performance once real variation from shifts, batches and tooling wear is present.
What if the machine fails the capability target?
Establish first whether the failure is spread or centring. Off-centre results are often correctable by adjustment or tooling; excessive spread usually points to the machine, the fixture or the measurement system, and is the supplier's responsibility if capability is a contractual acceptance criterion.
Who pays for a re-run?
Where the RFQ makes capability an acceptance criterion, re-testing after a failed study is normally at the supplier's cost. If no criterion was specified, the buyer typically has no contractual basis to demand it — which is why the clause belongs in the enquiry, not the punch list.
Which characteristics should be studied?
Only the ones that matter commercially or for safety — typically three to eight critical-to-quality characteristics named in the specification. Studying every dimension inflates cost and dilutes attention on the ones that drive rejects.
Procurement answers
Capability requirements in a machinery RFQ
Capability indices are a contractual acceptance tool, not just a statistics exercise — they belong in the RFQ.
How are capability requirements written into an equipment purchase?
State the characteristic to be controlled, the tolerance, the index and target value, the sample size, the material and operating conditions for the trial, and whether the study is run at the supplier works, on site, or both. Tying acceptance and the final payment milestone to a demonstrated result removes most of the argument at commissioning.
What information is needed for an industrial RFQ?
A comparable RFQ states project location, the process or product being made, required capacity or throughput, technical and utility constraints, quality or certification requirements, delivery terms, installation and commissioning scope, target timeline, indicative budget band and whether financing is required. Without those fields, suppliers quote different scopes and the offers cannot be compared line by line.
How does supplier matching work at Global B2B Group?
A buyer submits a requirement, it is structured into a procurement request, relevant supplier categories are identified, and sourcing options are reviewed against the stated technical and commercial constraints. The buyer then continues discussions directly with the suppliers or partners that fit. Global B2B Group is supplier-neutral and does not manufacture equipment or represent a single manufacturer.
Acceptance criteria and capability targets are captured in the structured RFQ before suppliers quote.
Continue on the platform
- quality assurance plan (QAP)
How capability evidence is scheduled and signed off.
- factory acceptance test scope
Witnessed testing before shipment.
- industrial machinery categories
Equipment categories, industry clusters and geo markets.
Global B2B Group is supplier-neutral: we do not manufacture equipment or represent a single manufacturer. We are not a bank or lender — financing options may be explored with external financing partners, subject to eligibility, due diligence and lender approval.
Prepare this project before contacting manufacturers
Scope this page helps prepare: Machine and process capability requirements that belong in the RFQ and in the acceptance test, not in a post-delivery argument.
Who it is for: Quality, engineering and procurement teams specifying new machinery or a production line.
Confirm before manufacturers are contacted: product and capacity, process and equipment scope, site and utilities, automation and quality requirements, budget range and timeline, and which figures are still assumptions.
Next step: Define the measurement method, sample period, tolerances and out-of-range behaviour you will accept at FAT and SAT.
- Initial conversation
- Structured requirements
- Project or RFQ brief
- Human review
- Appropriate routing, where justified
Industrial RFQ-readiness checklist
- Product and target market
What is produced, for which market, and the product or formulation specification.
- Capacity, throughput and shifts
Units per hour, per day and per year, shift model, planned growth.
- Inputs and outputs
Raw materials, input variability, packaging formats, output specification.
- Process steps and equipment scope
Process sequence, machine scope, required interfaces to existing lines.
- Automation, quality, traceability, safety
Automation level, control system, quality checks, traceability and safety standards.
- Site, building, utilities, environment
Site status, building dimensions, electricity, water, gas, steam, refrigeration, drainage, environmental limits.
- Installation, commissioning, training, service
Who installs, who commissions, training scope, spare parts and maintenance expectations.
- Budget, financing, contribution, timeline
Budget range, financing need, buyer contribution, required delivery and start-up dates.
- Compliance, approvals, tender constraints
Certification, permits, internal approvals, public-procurement or tender rules where relevant.
- Facts, assumptions and open questions
What is confirmed, what is assumed, what still needs engineering or supplier confirmation.
- Acceptance-test criteria
FAT and SAT scope, measurement method, sample period, tolerances, downtime and out-of-range behaviour.
What information do manufacturers need before reviewing a machinery project?
Product and capacity, process steps and equipment scope, site and utility conditions, automation and quality requirements, installation and commissioning expectations, budget range and timeline, plus a clear separation between confirmed facts and assumptions.
What is the difference between a machinery quote and a project brief?
A quote prices a defined scope. A project brief defines the scope: objective, capacity, process, interfaces, site conditions, acceptance criteria and open questions. Without the brief, quotes are not comparable.
Does Global B2B Group connect buyers directly to manufacturers?
No. Buyers are never automatically connected to a manufacturer or supplier. David and the human team review qualifying projects before any appropriate external routing.
What is the minimum project size for human review?
Formal project review, supplier routing and financing-pathway discussion generally start from an expected total project value of USD 250,000. Below that, planning guidance and calculators remain available.
Indicative planning support only. Global B2B Group is not a machinery manufacturer, engineering or EPC contractor, bank, lender, credit provider or pricing authority, and does not guarantee a manufacturer, equipment availability, technical performance, factory acceptance, final price, delivery, installation, financing or project outcome. WhatsApp is a technology-assisted, human-reviewed way to start or continue a project conversation.
Put capability acceptance into the RFQ, not the punch list
Send the characteristics that matter, the tolerances and the capability level you need. We package them into a structured enquiry so competing machine builders quote against the same acceptance criteria — free for buyers.
Build an equipment RFQGo deeper on the platform
Related terms
Cycle Time
Cycle time is the elapsed time required for a machine or process step to complete one full repeatable operation, from the start of one unit to the start of the next.
Factory Acceptance Test (FAT)
Formal verification test conducted at the supplier's manufacturing facility to confirm equipment meets agreed specifications before shipment to site.
Overall Equipment Effectiveness (OEE)
Composite performance metric combining availability, performance and quality rates to express how effectively a production asset converts planned time into good output.
Site Acceptance Test (SAT)
Formal verification test performed after equipment installation at the buyer's site, confirming it operates correctly under actual site conditions and integrations.
Technical Specification
Structured document describing the functional, performance, materials, and interface requirements an item of equipment or system must meet.
Prototype and Pilot Line
Prototype and pilot line refers to a reduced-scale or pre-production version of a manufacturing process used to validate design, process parameters and product quality before committing to full-scale investment.
Throughput
Throughput is the actual quantity of product a machine, line or facility produces over a defined period, expressed as units, mass or volume per unit time.
More in Engineering & Custom Machinery
Reference content only. Global B2B Group is independent of equipment manufacturers and financing institutions; definitions are provided for education and do not constitute engineering, financial or legal advice.