Skip to content

Free template

Updated

I-MR chart template (individuals and moving range)

An I-MR chart (individuals and moving range chart, also called an XmR chart) is the control chart for one reading at a time: one viscosity per batch, one torque audit per shift, the changeover minutes each day. The individuals (I) chart plots each reading against a centre line and control limits. The moving range (MR) chart plots the difference between each reading and the one before, which is how the chart measures short-term variation when there are no subgroups. The limits are the mean ± 2.66 × the average moving range, and the moving range limit is 3.267 × the average moving range (NIST/SEMATECH e-Handbook of Statistical Methods, sections 6.3.2.1 and 6.3.2.2). Page 1 of the PDF has the header, 25 readings with their moving ranges, and both charts with lines for the limits and for 1σ and 2σ. Page 2 has the formulas and constants, five run rules, which chart fits which data, when to recalculate and what to do when a rule fires. Page 3 is the worked example drawn as both charts, with a reaction log. The Excel version takes up to 100 readings, sets the limits from a baseline you choose and keeps them frozen for later readings, flags five Western Electric rules on the I chart and points above the limit on the MR chart, starts a second phase with its own limits after a deliberate change, and draws both charts.

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

Page 1 of the I-MR chartHeader fields for process, characteristic, units, gauge and specification, boxes for the baseline mean, average moving range and sigma, a table for 25 readings with date, value, moving range and rule broken, and an individuals chart with lines for the limits and for 1 and 2 sigma above a moving range chart.
25 readings per printed sheet, 100 in ExcelMoving rangesI chart and MR chartLimits from a baselineFive run rulesSecond phase after a changeWorked exampleReaction log

When to use it

When to use an I-MR chart

  • When you get one measurement per period and no subgroups: a viscosity or pH per batch, a torque audit per shift, daily changeover minutes, scrap in kilograms per day.
  • When readings come slowly or each one costs a lot, such as a destructive test or a lab result per lot, so waiting for subgroups of 4 or 5 would hide a change for days.
  • For a number tracked once a day or week, to tell a real change from normal ups and downs before anyone reacts.
  • Not when you can measure small subgroups of parts made one after another (use an X-bar and R chart), and not for the share of units defective (a p chart) or counts of defects (a c or u chart).
  • Not for readings taken so close together that each one follows the last, such as an oven temperature every minute: the limits come out too narrow and the chart signals all the time.

How to fill it in

  1. 1

    Check the measurement

    The gauge or lab method has to see the variation you want to chart. A gauge R&R study answers that before you start.

  2. 2

    Record readings in time order

    One row per batch, shift or day, with the date or batch number and a note of anything that changed. The moving range is the difference from the reading before, ignoring the sign.

  3. 3

    Set the limits from a baseline

    Take 20 or more readings from normal running. Centre line = their mean; MR̄ = the mean of their moving ranges. I chart limits = mean ± 2.66 × MR̄; MR chart upper limit = 3.267 × MR̄.

  4. 4

    Freeze the limits

    Judge every later reading against the same limits. Treat limits from fewer than about 100 readings as trial limits and update them once, if nothing in the process has changed.

  5. 5

    React to the rules

    A point beyond a limit, 2 of 3 beyond 2σ, 4 of 5 beyond 1σ, 8 in a row on one side or 6 in a row rising or falling. Find what was different for that reading, fix it and log it.

  6. 6

    Start a new phase after a deliberate change

    A new machine, material or method gets its own baseline and limits. Do not recalculate because the latest points look different: catching that is the chart's job.

Run rules

The five rules the Excel version flags

The Western Electric rules as listed in the NIST/SEMATECH e-Handbook, section 6.3.2, with σ = MR̄ ÷ 1.128. Rules 2 to 5 can be switched off: each one catches smaller shifts and raises false alarms, from about one in 371 points with rule 1 alone to about one in 92 with all of them (NIST).

  • 1

    Signal
    One point beyond a control limit
    Causes to check first
    A one-off: a wrong setting, a different material lot, a measurement or typing error
  • 2

    Signal
    2 of 3 points in a row more than 2σ from the centre line, same side
    Causes to check first
    The average starting to shift
  • 3

    Signal
    4 of 5 points in a row more than 1σ from the centre line, same side
    Causes to check first
    A small shift that has held for several readings
  • 4

    Signal
    8 points in a row on one side of the centre line
    Causes to check first
    The average has moved: a new lot, setting, supplier or person
  • 5

    Signal
    6 points in a row, each higher than the last (or each lower)
    Causes to check first
    Drift: wear, build-up, a material that changes slowly
  • MR

    Signal
    A moving range above 3.267 × MR̄ (the only rule on the MR chart)
    Causes to check first
    A sudden jump between two readings

A filled-in example

Illustrative, not a benchmark

An example: the viscosity of a water-based coating, one reading per batch, specification 380 to 450 cP (illustrative numbers). The Excel version's Example sheet has the same 30 readings.

The example's 30 batches on the I chart and the MR chart

Individuals (I) chart, viscosity in cP

Moving range (MR) chart, cP

I chart: mean ± 2.66 × MR̄MR chart: UCL = 3.267 × MR̄

MR̄ is the average of the 19 moving ranges in the baseline: 182 ÷ 19.

  1. UCL =412.0 + 2.66 × 9.58= 437.5 cP
  2. LCL =412.0 − 2.66 × 9.58= 386.5 cP
  3. MR UCL =3.267 × 9.58= 31.3 cP

Individuals chart of 30 viscosity readings, one per batch, with limits from batches 1 to 20: mean 412.0 cP, UCL 437.5 cP, LCL 386.5 cP. Batch 24, at 441 cP, is the only point outside the limits, although it is inside the 450 cP specification. Moving range chart: MR-bar 9.58 cP and UCL 31.3 cP; batch 24's moving ranges, 20 and 19, are under it.

The limits come from the process, not from the 380 to 450 cP specification, which is why the chart flagged a batch the specification check passed. Batch 24 is outside the baseline, so the limits stay at 386.5 to 437.5 cP.
  • Baseline, batches 1 to 20: mean = 8,240 ÷ 20 = 412.0 cP. The 19 moving ranges add to 182, so MR̄ = 9.58 cP.
  • I chart: 2.66 × 9.58 = 25.48, so UCL = 437.5 cP and LCL = 386.5 cP. MR chart: UCL = 3.267 × 9.58 = 31.3 cP.
  • Batch 24 reads 441 cP: above the UCL (rule 1), although it is inside the 450 cP specification. Its moving ranges, 20 and 19, stay under 31.3.
  • The thinner flow meter had not been reset after a line flush, so batch 24 got less thinner. No other rule fires in the 30 batches.

The chart flagged a batch that the specification check passed. The meter reset was added to the flush checklist and logged. Batch 24 is outside the baseline, so the limits stay at 386.5 to 437.5 cP.

Common mistakes

  • Using the specification as control limits

    Specification limits say what the customer accepts; control limits say what the process does. Batch 24 in the example passed the specification and still showed a change worth finding.

  • Limits computed from the data you are judging

    Recalculating the limits each time new points come in lets a shift drag the limits along with it, so it never shows. Set them from a baseline, then freeze them.

  • Reacting to every point

    A reading higher than the last one is not a signal. Adjusting the process for points inside the limits with no rule broken chases normal variation; act on the rules and leave the rest.

  • Trusting the rules on skewed or linked readings

    The zone rules assume roughly normal readings. Strongly skewed data, such as time between failures, gives more false alarms, and readings so close together that each follows the last make the limits too narrow. Look at a histogram and the time order first.

Download the template

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

Run this template in LeanSuite

In LeanSuite quality checks run as digital checklists with number limits, an out-of-spec answer can raise a tag or start a CAPA automatically, and KPI Builder tracks quality KPIs against targets.

See it in a 45-minute demo

FAQ

I-MR chart template (individuals and moving range): common questions

More free templates

All templates
Browse all 189 templates

Can't find the template you need?

Tell us what you're looking for. We'll make it and send it to your inbox, free.

A shift handover sheet, a red tag log, a changeover checklist for one machine: if your team uses it on the floor, ask.

By sending this form you agree to our Terms of Service and Privacy Policy.

Pass it on

We want LeanSuite to be the best place on the internet for lean help. If this was useful, send it to someone on your team or in your network who needs it.