muono
Manufacturing

The line ran. Nobody agrees where the hour went.

Changeover is the shift's biggest single loss and the least well recorded. The plan says forty minutes, the line says ninety, and the difference is remembered rather than measured.

Output against line rateOne line, one shift
RATED
THE MISSING HOUR
ACCOUNTED FOR BELOW, MINUTE BY MINUTE
06:0014:00
Line PLC
Changeover log
Quality
Maintenance
The line
Every station, its cycle time and the condition it was in when the run started.
The changeover
Purge, tool change, heat, first-off, each step timed against the plan and against the last crew that ran it.
The output
Scrap and rework booked to the station that produced them, not to the shift that noticed.
The derivation

Built line by line, from your own systems.

Nothing here is inferred. Each line names the system it came from, so the shift lead can check a number before handover and the plant manager can defend it at the weekly review. Values fill in from your own line.

Step
What it establishes
Reading
Actual line rate against rated speed, minute by minute.
Line PLC · counter and speed tags
Record
The changeover in progress and the work open on the line.
Changeover log · maintenance work order
Calculation
Lost units, taken as the area between rated and actual.
Split by stop reason, with each stop shown
Source
The shift report the figure lands in, and the quality record beside it.
Shift report · quality system
The same derivation runs for every line, every shift, without anyone assembling it.
The line

Where the shift's output actually went.

Counter, speed, stop reasons and the work open on the line sit in four systems. The shift report is written from memory because assembling the real answer takes longer than the shift.

CV-2 · CONVEYOR
RUNNING
FL-1 · FILLER
PE-4
PE-5
M-3 · DRIVE
SI-1
Line output
4 sources
PLC · stop log · quality · SAP PM

A line that misses its number has usually lost the time in small pieces: a jam, a slow-feed, a five-minute clean nobody logged. Micro-stops are the ones that never make the log, and on most lines they are the largest single availability loss. Hold each stop with its duration, its reason and the station it happened on, and the shift report is read off the line instead of written up afterwards.

It also settles the argument about rate. Rated speed, actual speed and the count that came off the end are three different numbers on most lines, which is why the performance term in OEE is the one nobody trusts. Held together, the three become one calculation you can open.

Changeover

Forty minutes planned, ninety taken.

The plan lives in scheduling, the reality in the operator's head, and the difference in nobody's system. Changeover is the largest recurring loss on most lines and the least examined.

EX-1 · EXTRUDER
HOPPER
Z1
Z2
Z3
DIE-4
TAKE-OFF
TI-9
Changeover time
3 sources
PLC · changeover log · MES

A changeover is a sequence: purge, cool, tool change, heat, first-off, sign-off. Each step has a duration somebody could take minutes out of, but only if the durations exist in the first place. Muono records the sequence against the machine and the product pair, so the ninety minutes has a shape and not just a total. That is the input a SMED exercise needs and rarely gets.

The comparison that matters is between crews and between products, not against a plan written once. When each changeover carries its own record, the fast ones become instructions instead of anecdotes.

Quality

First-pass yield, traced to a cause.

A reject rate is a number without a cause attached. What the line needs is the batch, the setting and the reading that were live when the rejects started.

CV-3
VI-2 · INSPECTION
INSPECTED
DV-1
REJECT
ACCEPT
QI-1
First-pass yield
4 sources
Inspection · PLC · batch · MES

Inspection gives you a verdict. It does not give you a cause. Muono holds the reject against the batch, the machine settings in force at the time and the process readings from that moment, so a drift in first-pass yield points at a die temperature or a tool change and not at the shift that reported it.

It is the same structure as the rest of the plant, read in the language of quality. When a customer raises a complaint and you owe them an 8D, the batch, the settings and the readings behind that batch are already together.

The pilot

One line, five stages, a figure at the end.

A paid pilot on a single line, then an annual licence per site. We take the baseline before anything changes, because a changeover time you did not measure beforehand is not a result.

  1. Scope
    Pick the line and the loss it has to explain.
  2. Connect
    Line PLC, MES, maintenance system and the quality records beside them.
  3. Baseline
    Measure how the line runs today, before anything changes.
  4. First answers
    Your team asks real questions and checks the sources behind them.
  5. Review
    Compare against the baseline and decide whether it widens.
What it is worth

Put your own numbers in.

We will not publish a result we have not measured. So here is the arithmetic instead, on your figures rather than ours. Change any of them.

The working
14 a week x 52 x 30 min x £2,200/h
A year, on this line
£801k

This is the size of your problem, not a claim about our effect on it. Muono does not take a minute out of a changeover. It records the sequence step by step so the fast crews become a standard instead of an anecdote, and how much of the overrun you close is yours to judge.

If nobody can tell you the average overrun on this line today, that is itself the finding.

Your questions, answered

Do you connect to our MES and historian?
Yes, read-only, on a schedule you set. Station tags are mapped once to the line model. After that new cycles land on the right station without anyone mapping them again.
We already measure OEE. What changes?
OEE tells you the size of the loss. This tells you which step produced it and what was on the record at the time. That is the part that lets someone fix it rather than report it.
Can it read our work instructions and engineering changes?
Documents are read in and attached to the station or tool they govern, so the version in force during a run is the version you see when you review that run.
Does this need new sensors on the line?
Usually not to start. Most of what a changeover analysis needs is already in the MES, the historian and the inspection records. It is just held apart.

Start with one line and one changeover.

Request a pilotBook a walkthrough