Six Sigma Software

Run Lean Six Sigma projects efficiently

With numiqo, you can use statistical analysis to improve quality and processes, identify sources of variation, and make informed decisions.

Selected customers

numiqo helps you run Lean Six Sigma projects efficiently, from measurement system analysis to process capability, SPC and designed experiments. The software focuses on the core statistical tools quality teams need without forcing users into a complex desktop statistics package.

If you are comparing Six Sigma software with established tools, our Minitab alternative page gives you a practical overview of numiqo for SPC, DoE, MSA and process capability.

Software for Lean Six Sigma projects

Six Sigma projects are built on measurable process improvement. You need clear metrics, reliable analyses and understandable charts. numiqo combines these elements in a browser-based workflow: import data, choose the right analysis and interpret the results.

Statistical support for DMAIC

DMAIC stands for Define, Measure, Analyze, Improve and Control. numiqo supports the statistical work in each phase; your team sets the project goals, assigns actions and manages implementation.

What quality management software means here

numiqo provides statistical quality analysis: evaluating measurements, investigating variation and checking process performance. It is not a full quality management system (QMS) for document control, audits or corrective and preventive action (CAPA). The workflow described here uses imported or pasted data, rather than automatic data collection from factory equipment. Your team can use the statistical results in its existing quality system and control plan.

Important Six Sigma analyses

For typical Lean Six Sigma projects, numiqo provides the key statistical tools used in quality and process improvement:

Measurement System Analysis (Gage R&R)

Check whether your measurement system is reliable enough before evaluating process data. Measurement system analysis shows repeatability, reproducibility and the share of variation caused by the measurement process.

Process Capability Analysis

Calculate Cp, Cpk, Pp and Ppk to assess whether a process can meet specification limits consistently. Histograms and capability indices help you evaluate process performance quickly.

Control Charts and SPC

Monitor processes with control charts and detect trends, shifts or outliers early. SPC helps you review measurements collected over time.

Design of Experiments

Plan experiments systematically and identify the factors that truly influence your process. Full factorial, fractional factorial and other DoE designs support the Improve phase of DMAIC.

Monte Carlo Simulation

Model uncertainty with input distributions and output equations. Monte Carlo simulation helps Six Sigma teams estimate risk, specification-limit failures, PPM, and sensitivity when process inputs vary.

Predictive Analytics

Build decision tree and Random Forest models to classify outcomes, estimate future values and identify the variables that drive process behaviour. Predictive analytics helps Six Sigma teams move from explaining past variation to forecasting likely outcomes.

Worked example: investigating shaft diameter

A machining team wants to center shaft diameters on 20.000 mm, with specification limits of 19.950 and 20.050 mm. Its question is whether a change in spindle speed and feed can bring the average closer to target. All data below are simulated for learning, not customer results. The machine settings are illustrative, not production recommendations.

Use the links below to load the simulated data directly into numiqo and analyze the examples for free. Then select the variables and enter the settings described in each step. Keep the supplied row order and treat Diameter_mm as a metric variable. Results below are rounded.

1. Check the gage

Load the 50 simulated gage readings. These represent one operator repeatedly measuring a reference part of known diameter 20.000 mm. In the MSA calculator, select Continuous data and Crossed, then select only Diameter_mm as the measurement; leave parts and operators unselected for a Type 1 study. Set reference = 20, tolerance range = 0.100, study variation = 6 and percentage of tolerance = 20.

Result: Mean = 20.000254 mm, standard deviation = 0.000871 mm, Cg = 3.83 and Cgk = 3.73. Repeatability and bias are small relative to the stated tolerance in this exercise. Next decision: proceed with the same operator and gage for the example. For a production study involving several operators and parts, also assess reproducibility with a Gage R&R study; this Type 1 check does not cover it.

2. Inspect the baseline in time order

Load the 40 simulated baseline measurements in SPC. In the I-MR calculator, select Diameter_mm, use a moving range of 2 and inspect the test for points beyond three standard deviations. Each row represents a successive shaft.

Result: The I-chart center line is 20.02196 mm, with limits of approximately 19.99132 and 20.05260 mm. The moving-range upper limit is 0.03764 mm. No points exceed these limits. Next decision: review other run-pattern tests and the process history before treating the process as stable. Passing this one check on 40 measurements is not proof of long-term stability. Control limits describe process behavior; they are not the specification limits.

3. Compare the baseline with specifications

Load the same baseline data in the process capability calculator. Select normal analysis, Diameter_mm, LSL = 19.950, USL = 20.050, target = 20.000, subgroup size = 1 and moving range length = 2. This exercise uses simulated normal measurements; with real data, check stability and distribution suitability first.

Result: Cp = 1.63, Cpk = 0.92 and Ppk = 0.95. Although all 40 readings are within specification, the mean is close enough to the upper limit to reduce Cpk substantially below Cp. Next decision: investigate settings that center the process, rather than declaring it satisfactory from the observed pass count.

4. Test candidate settings with DoE

Load the simulated 12-run experiment in DoE. The plan is a two-factor full factorial design: spindle speed at 1000 and 1400 rpm, feed at 0.10 and 0.20 mm/rev, with three independent runs at each combination. The data are already in randomized run order. To reproduce the design structure in DoE, use two numeric factors with those levels, three replicates and no center points or blocks. A newly randomized plan may have a different order; keep the supplied responses paired with their original factor settings.

For the loaded example, open Analyse, select Speed_rpm and Feed_mm_per_rev as factors, Diameter_mm as the response, and include their two-way interaction. Do not select Run_order as a factor. Result: Increasing speed lowers the fitted mean by 0.020 mm; increasing feed raises it by 0.020 mm. The fitted interaction is zero in this constructed dataset. The model is diameter = 20.019 - 0.010A + 0.010B mm, where A = (speed - 1200)/200 and B = (feed - 0.15)/0.05.

Next decision: try 1400 rpm and 0.10 mm/rev in a confirmation run: the predicted mean is 19.999 mm, close to the 20.000 mm target. This experiment models the mean response; it does not establish reduced variation or performance outside the tested settings.

5. Review subsequent measurements

Load the 40 simulated follow-up measurements in SPC, then load them in process capability. These illustrate a separate confirmation sample at the candidate settings. Repeat the I-MR and normal capability analyses with the same specifications and estimation settings, analyzing this sample separately from the baseline.

Result: Mean = 20.00227 mm, Cpk = 1.74 and Ppk = 1.86. The provisional I-chart limits are 19.97477 and 20.02976 mm; the moving-range upper limit is 0.03378 mm. No points exceed these limits. The sample is closer to target and has better estimated capability. Next decision: collect measurements across shifts, batches and tool wear before accepting sustained improvement. Once a stable reference period is established, fix the control limits for subsequent monitoring and agree who investigates a signal; do not recalculate limits merely to absorb an unfavorable change.

For the statistical background on stability and distribution assumptions, see the NIST guide to process capability.

Why use numiqo for Six Sigma?

• Fast start: Analyses are available directly in the browser and guide users to the right statistical output.
• Relevant methods: MSA, process capability, control charts, hypothesis tests, DoE, Monte Carlo simulation, and predictive analytics cover central Lean Six Sigma requirements.
• Data privacy: Statistical calculations run locally in your browser without uploading your dataset. Optional AI Help sends the content needed for your request to numiqo's AI server.
• Fair pricing: numiqo is a cost-effective Six Sigma software solution for individual users, teams and organisations.

If you want to give more employees access to serious statistics without Minitab-level seat costs, numiqo is the right choice.

Efficiently move from problem to improvement

Whether you need root cause analysis, process capability evidence or ongoing process monitoring, numiqo helps you carry out Lean Six Sigma projects efficiently and make decisions based on data.

Start with an analysis: calculate process capability, run a measurement system analysis, create a control chart or build a predictive analytics model.