Book Demo →

KaizenFlow AI Certified Plant Manager — Course

Module 01 · 24 min · 2 fig.

The plant on one page

As a plant manager you don't watch one line — you run the whole floor against a number and a ledger. This module teaches you to read rolled-up OEE/TEEP and the savings ledger as a single operating view, so you spend attention where it changes money.

Roll-up, don't average blindly

OEE rolls up from station to line to plant, but a plant-level OEE number hides where the loss lives. Your job is to read the roll-up top-down and drill on the worst contributor, not to celebrate or panic over the headline. A high plant OEE can still hide one constraint line bleeding the schedule.

Read OEE as its three factors — Availability, Performance, Quality — at every level. The factor that's lowest tells you which of the Six Big Losses to chase. A plant strong on Quality but weak on Availability has a downtime problem, not a scrap problem, and that distinction decides who you call into the room.

  • Availability low → downtime/changeover losses
  • Performance low → speed/minor-stop losses
  • Quality low → defect/startup-reject losses
FIG. 1.A THE ROLL-UP, READ TOP-DOWN KF·PM-01
Plant82%HEADLINELine 184%OEELine 286%OEELine 368%CONSTRAINT · A 71%Line 485%OEE
A healthy 82 headline with the constraint bleeding at 68. The roll-up isn't for celebrating or panicking — it's for finding the worst contributor and drilling into its weakest factor, here Line 3's availability at 71%.

OEE tells you 'how well,' TEEP tells you 'how much is left'

OEE measures how well you run during scheduled time. TEEP measures the same effectiveness against all calendar time, so it exposes unscheduled hours — the capacity you own but aren't using. On one page, OEE drives improvement work and TEEP drives capacity and capex conversations.

When someone says the plant is 'maxed out,' the TEEP number settles it. If TEEP is well below OEE, you have idle calendar time before you have a true constraint.

The ledger is the second half of the page

Effectiveness numbers say where you're losing; the verified-savings ledger says what you've actually recovered and banked. Reading them together turns a status dashboard into an operating view: losses on the left, finance-verified recoveries on the right. If OEE is improving but the ledger isn't moving, your gains aren't reaching the P&L — and that gap is itself a signal to investigate.

FIG. 1.B ONE PAGE, TWO HALVES KF·PM-01
EFFECTIVENESS VIEWOEE 82%where you're losingSAVINGS LEDGER$38,400finance-verified this quarterOne pageLOSSES LEFT · BANKED RIGHT
Effectiveness numbers name the losses; the ledger proves what you recovered. When OEE climbs and the ledger doesn't move, the page is telling you your gains never reached the P&L — and that gap is your next question.
Key takeaway

Read top-down: OEE shows how well you run, TEEP shows how much capacity is left, and the verified ledger shows what actually hit the P&L. One page, three lenses.

Module quiz · question 1 of 3

Plant OEE is up two points week-over-week and the headline tile looks healthy. But the savings ledger is flat, and Line 3 — your constraint — shows Availability as its weakest factor while two strong lines carry the average up.

Q1What do you do first?

Module quiz · question 2 of 3

Q2Plant TEEP is far below plant OEE. What is that gap primarily telling you?

Module quiz · question 3 of 3

The plant roll-up shows Quality strong at every level, Performance middling, and Availability the weakest factor on the two biggest lines.

Q3Which loss family do you chase, and who do you call into the room?

Module 02 · 24 min · 3 fig.

Running the operating rhythm

A live dashboard only earns its keep if it drives a rhythm. This module shows how to wire KaizenFlow into a tiered cadence — shift huddle to daily tier board to weekly to monthly ops review — so the same data escalates cleanly instead of being re-litigated at every level.

Tier the conversation, escalate by exception

Each tier reads the same platform at a different altitude. The shift huddle works live tiles and reason codes; the daily tier board reviews yesterday's losses and open actions; the weekly review looks at trends and the action backlog; the monthly ops review works rolled-up OEE/TEEP and the ledger. Lower tiers resolve what they can and escalate only the exceptions they can't.

The discipline is escalation by exception. If a downtime cause is recurring and a shift can't fix it, it rises to the next tier with its reason-code history attached — not as a fresh complaint, but as evidence.

FIG. 2.A THE TIERED OPERATING RHYTHM KF·PM-02
Shift huddleLIVE TILES · REASONSDaily tier boardYESTERDAY'S LOSSESWeekly reviewTRENDS · BACKLOGMonthly ops reviewOEE/TEEP · LEDGERONE PLATFORM
Four tiers, one platform, no re-litigation. Each tier reads the same data at its own altitude, resolves what it can, and passes up only the exceptions it can't — with the reason-code history attached as evidence.

Let the platform set the agenda

Don't build slides. Each meeting opens on a fixed view: live line view and reason logging for the huddle, the ranked next-best-action list and anomaly alerts for daily/weekly, rolled-up OEE/TEEP plus the savings ledger for monthly. The agenda is the screen.

This kills the two failure modes of plant meetings — debating whose number is right, and rediscovering problems. When the reason codes and the ranked actions are on the wall, the meeting is about decisions and owners, not data archaeology.

FIG. 2.B THE AGENDA IS THE SCREEN KF·PM-02
HUDDLEDAILYWEEKLYMONTHLYLive line view + loggingRanked actions + alertsLoss trends + backlogOEE/TEEP roll-up + ledgerOPENING VIEWIN SUPPORTNOT THIS TIER
Each meeting opens on its fixed view — nobody builds slides, nobody debates whose number is right, and the meeting time goes to decisions and owners instead of data archaeology.

Close the loop or the rhythm dies

A cadence without follow-through becomes theater. Every action raised gets an owner and a due date in the platform, and every review opens by checking last period's actions before looking at anything new. The anomaly alerts and failure forecasts feed the agenda so you're acting ahead of breakdowns, not just reporting them after.

FIG. 2.C CLOSE THE LOOP OR THE RHYTHM DIES KF·PM-02
Action raisedIN REVIEW01Owner + due dateIN THE PLATFORM02Work the actionBETWEEN REVIEWS03Review opens on itBEFORE ANYTHING NEW04CLOSED OR CARRIED
Every review opens by checking last period's actions before anything new. An action either closes or carries forward in the open — that loop is what keeps the cadence from becoming theater.
Key takeaway

The platform is the agenda. Tier the conversation, escalate by exception with the reason-code history attached, and open every review by closing the last one's actions.

Module quiz · question 1 of 3

The same changeover-related downtime cause has shown up in the shift huddle's reason codes four days running. The shift teams keep logging it and working around it, but the fix is outside what they can do on their own.

Q1What happens on day five?

Module quiz · question 2 of 3

Q2Every tiered review — daily, weekly, monthly — opens the same way. With what?

Module quiz · question 3 of 3

Your monthly ops review burns its first twenty minutes the same way every month: two department heads comparing their own spreadsheets to argue about whose downtime number is right.

Q3What's the structural fix?

Module 03 · 24 min · 2 fig.

Allocating the improvement budget

You have finite engineering hours and a finite improvement budget. This module is about spending both against ranked, verified opportunity instead of the loudest voice or the newest fire.

Rank by recoverable money, on the constraint first

The platform produces an AI-ranked next-best-action list tied to the Six Big Losses and the savings each would recover. Your allocation default is to fund top-down from that list — but weighted toward the constraint. An hour saved on a non-constraint line often banks nothing, because the schedule was never waiting on it.

So read two things together: the ranked recoverable savings, and whether the opportunity sits on the line that actually gates throughput. The best spend is high recoverable value on the constraint.

FIG. 3.A RANKED DOLLARS VS THE CONSTRAINT KF·PM-03
01Slow cycles — Line 1 filler$5,200/wkNON-CONSTRAINT · SLACK02Changeover — Line 3 press$4,100/wkCONSTRAINT · GATES THROUGHPUT03Micro-stops — Line 2 labeler$1,800/wkNON-CONSTRAINT
The biggest raw number sits on a line the schedule was never waiting on. Recoverable dollars rank the list; the constraint flag decides where the money actually banks.

Verified beats estimated

An estimated saving is a hypothesis; a verified saving is a fact the finance team will stand behind. When you allocate the next dollar, prefer opportunities whose comparable past actions actually verified in the ledger over ones with big estimates and no track record. The ledger is your evidence that this class of fix converts.

This also disciplines vendors and internal pet projects: if a proposed improvement can't be measured into the ledger, you can't prove it worked, and you shouldn't treat its savings as real.

FIG. 3.B FUND THE FACT, NOT THE HOPE KF·PM-03
BIG ESTIMATE$9,000/mono comparable verified historyVERIFIED CLASS$5,500/mo3 comparable fixes in the ledgerFund the factHYPOTHESIS VS EVIDENCE
An estimate is a hypothesis; a verified class of fix is a fact finance will stand behind. When the next dollar is on the line, the smaller proven number usually beats the bigger hope.

Budget is a portfolio, not a queue

Don't drain the whole budget into one big bet. Mix a few high-confidence, fast-verifying actions to keep the ledger moving and credibility high, with one or two larger structural fixes on the constraint. Reassess every cycle as the ranking updates — the next-best action changes as losses shift, so last month's plan is not this month's.

Key takeaway

Fund top-down from the ranked list, weighted to the constraint and toward fixes whose savings actually verify. Estimated value is a hypothesis; ledger value is proof.

Module quiz · question 1 of 3

Two items sit near the top of the ranked list. A: a large estimated saving on a non-constraint line, no comparable verified history. B: a smaller saving on your constraint line, from a class of fix that has verified in the ledger three times.

Q1Where does the next engineering hour go?

Module quiz · question 2 of 3

Q2Why does an opportunity with verified comparable savings beat one with only a large estimate?

Module quiz · question 3 of 3

This quarter's improvement budget just covers the #1 ranked item: a structural constraint fix that would consume all of it and take two quarters to verify. Items #2–#5 are smaller, high-confidence fixes that verify within weeks.

Q3How do you allocate?

Module 04 · 22 min · 3 fig.

Capacity decisions with TEEP

Capacity requests are where plants spend the most money on the least evidence. This module is about using TEEP to tell a real constraint from idle time before you sign a capex request for more of it.

TEEP is your capex gatekeeper

TEEP measures effectiveness against all calendar time — every hour you own, not just scheduled ones. Before approving capacity spend, read TEEP on the line in question. If TEEP is low, you have unused calendar time, and the cheapest 'new capacity' is the hours you already own. Buying equipment to solve idle time is paying for a problem you don't have.

The order is fixed: prove the line is genuinely constrained before you fund more of it. TEEP near the ceiling with demand still unmet is the signature of a real constraint.

FIG. 4.A CALENDAR TIME TO TEEP KF·PM-04
100%CALENDAR TIME−30%UNSCHEDULED−10.5%DOWNTIME−8.9%SLOW RUNNING−5.1%QUALITY45.5%TEEP
OEE starts at the 70-point mark and explains only the last three cuts; TEEP charges the line for the whole calendar. The biggest single loss here never touched a machine — it's the 30 points you didn't schedule.

Separate the kinds of 'not running'

A line that isn't producing is either unscheduled (idle calendar time), scheduled-but-down (an Availability loss), or scheduled-but-slow (a Performance loss). Each points to a different lever. Idle time means schedule more or sell the capacity; downtime means reliability and changeover work; slow running means speed and minor-stop work — none of which require capex.

Use the OEE-vs-TEEP gap to size how much capacity is hiding inside the asset you already own before you assume you need another one.

FIG. 4.B THREE KINDS OF NOT RUNNING KF·PM-04
30%IDLE —UNSCHEDULEDschedule it or sell it10.5%DOWN —AVAILABILITYreliability + changeover work8.9%SLOW —PERFORMANCEspeed + minor-stop work
Three kinds of 'not running,' three different levers — and none of them is a purchase order. The biggest bucket here is fixed with a schedule, not a machine.
Worked example — TEEP on one line, one week

The line exists for 168 hours this week. You scheduled it for 118 of them — utilization 70%. While scheduled, it ran at Availability 85%, Performance 85%, Quality 90%.

Utilization = 118 h ÷ 168 h ≈ 0.70 OEE = 0.85 × 0.85 × 0.90 = 0.65 TEEP = Utilization × OEE = 0.70 × 0.65 = 0.455 → 45.5%

Read the gap: OEE says the line runs at 65% while scheduled. TEEP says that against every hour you own, it delivers 45.5%. Those ~20 points are the unscheduled hours — capacity you already paid for, and the first place to look before any capex request.

Exhaust recovered capacity before bought capacity

Run the sequence: recover idle calendar time, then recover OEE losses on the constraint, then — only if TEEP is genuinely near its ceiling and demand persists — request capex. Funding equipment while OEE-recoverable hours sit on the floor buys you an expensive asset that will run at the same loss rate as the one you already have.

FIG. 4.C THE CAPEX ASK VS THE HOURS YOU OWN KF·PM-04
THE CAPEX ASK$1.4Msecond machine · 9-mo leadHOURS YOU OWN50 h/wkunscheduled calendar · same lineCheck TEEP firstRECOVERED BEFORE BOUGHT
Before the $1.4M signature, the TEEP read: 30% of the line's calendar — about 50 hours a week — is unscheduled. Exhaust the hours you already own, then the OEE losses on the constraint, and only then talk equipment.
Key takeaway

Don't buy capacity you already own. Low TEEP means idle time, not a constraint — exhaust recovered hours and OEE losses before any capex request.

Module quiz · question 1 of 3

A line lead wants a second machine because the line 'can't keep up with demand.' You pull the page: the line's TEEP is well below its OEE, with meaningful unscheduled calendar time and recoverable Availability losses.

Q1What's your call?

Module quiz · question 2 of 3

Q2A line shows a large gap between its OEE and its TEEP. What does that gap most directly represent?

Module quiz · question 3 of 3

Line 5 is scheduled around the clock apart from a short weekly maintenance window. TEEP reads 78% against OEE 81%, demand is still unmet, and the ranked list shows only small recoverable losses left on it.

Q3What is this the signature of?

Module 05 · 20 min · 2 fig.

Accountability that holds

A savings story is only worth what it can survive in an audit. This module is about the discipline that makes improvement stick: clear owners, real due dates, and a verified ledger that traces every claimed dollar back to a measured change.

Every action has one owner and a real date

Shared ownership is no ownership. Each opportunity you fund gets a single accountable owner and a committed due date in the platform, visible to the tier that raised it. The next-best-action list is where work originates; the owner-and-date discipline is what turns a ranked idea into a closed action.

Accountability that holds is reviewed in the open. Because the same actions appear at every tier of the rhythm, an owner can't quietly let a commitment slide — the board carries it forward until it's closed or consciously dropped.

The ledger makes the savings audit-proof

A claimed saving is only real once it's finance-verified in the ledger — tied to the action that produced it, the loss it addressed, and the measured before/after. That traceability is what lets you defend the plant's savings number to finance, to a corporate audit, or to your own board without hand-waving.

Hold the line on this: if a saving can't be measured into the ledger, it isn't counted. That rule protects the credibility of every number you do claim.

FIG. 5.A ANATOMY OF AN AUDIT-PROOF DOLLAR KF·PM-05
One verified dollarThe actionone ownerclosed on a dateThe lossSix Big Losses bucketthe line affectedThe measurementbaseline beforemeasured afterfinance sign-off
Trace any claimed dollar down this tree and it survives an audit: the action that produced it, the loss it addressed, and the measurement that proves it. If the measurement branch is missing, it isn't counted.

Drive the close rate, not just the open rate

It's easy to generate actions; what distinguishes a well-run plant is the rate at which actions close and verify. Track open vs. closed vs. verified as its own metric in the cadence. A growing pile of open actions with a thin verified ledger is a warning sign that the rhythm is generating activity without banking results.

FIG. 5.B OPEN, CLOSED, VERIFIED KF·PM-05
46OPENEDraised in reviews31CLOSEDwork done22VERIFIEDfinance-signed
Open actions measure activity; the verified bar measures results. A quarter where the first bar grows and the last one doesn't is a rhythm generating theater, not savings.
Key takeaway

One owner, one date, one ledger entry per claimed dollar. If it isn't finance-verified in the ledger, it doesn't count — that rule is what makes the savings story audit-proof.

Module quiz · question 1 of 3

Finance is reviewing the plant's quarterly savings before it rolls into the corporate number. One large item was logged 'completed' by a reliable engineer — but it has no before/after measurement in the ledger and no finance verification.

Q1What do you report?

Module quiz · question 2 of 3

Q2Which metric best shows your accountability rhythm is producing results rather than activity?

Module quiz · question 3 of 3

A funded action has been carried forward at three straight weekly reviews with no movement. It's assigned to 'the maintenance team,' and each week somebody different explains the delay.

Q3What's broken, and what's the fix?

Course complete.

You’ve worked all 5 modules. Sit the free practice exam to see if you’re ready for the real assessment — same format, instant grading, keyed back to the modules.

Take the practice exam → Get certified