The lead’s Kaizen course: manage the flow of the whole job, not the local task. Value stream mapping, lead-time vs touch-time, WIP limits, the constraint, parts pull, scheduling the yard, the Gemba walk, A3 root-cause, flow metrics and making gains stick — on ESI’s own repower and haul-out jobs, keeping the boatyard app, the ** marker and respect for people.
Before you start
Production Tech 2 is the lead’s Kaizen course. Production Tech 1 made every ESI employee waste-aware at their own station; this one teaches a lead to see and manage the whole job — the flow of a boat from request to delivery — because a boat torn apart for a week waiting on one part is the biggest waste we have. Value stream mapping, lead-time vs touch-time, WIP limits, the constraint, parts pull, scheduling the yard, the Gemba walk, A3 root-cause, flow metrics, and making gains stick — taught on ESI’s own repower and haul-out jobs. It keeps the Tech 1 spine: the boatyard app, the ** marker, and respect for people. A final exam unlocks your certificate.
Map work vs waiting; attack the wait, not the task.
Cap WIP, feed the constraint, pull the parts.
Gemba, A3, measure, standardise — respect for people throughout.
Chapter 1
Production Tech 1 taught you to see waste at your station. A lead sees the whole job.
This is the lead’s course. Production Tech 1 made every ESI employee waste-aware at their own station — 5S the bench, follow standard work, drop a ** note and tag the manager. That is the foundation and it never goes away. But the biggest waste at a boatyard is not at the bench. It is in the whole job: a boat sitting torn apart for a week while everyone is “busy” elsewhere.
| The local view (Tech 1) | The whole-job view (a lead, Tech 2) |
|---|---|
| Is my task efficient? | Is the boat moving toward delivery? |
| Am I busy? | Is the job waiting? |
| My station is clean and quick | The handoffs between stations are clean and quick |
| Measured in the hour | Measured in the boat’s days on the hard |
Here is the trap ESI lives with: a boat can have 40 hours of actual work in it and still take six weeks to deliver. Almost all of that calendar time is the boat waiting — for a part, for an approval, for the next trade, for a haul-out slot, for someone to finish the last 10%. Every one of those waiting days is pure waste, and a busy, clean bench does nothing to fix it.
Chapter 2
You cannot fix flow you cannot see. So draw it.
A value stream map is a simple picture of a whole job from customer request to delivered boat. You draw every step in order, mark how long each takes, and — the important part — mark the waiting between steps. It is the single most powerful tool a lead has, and it needs nothing but a whiteboard.
Touch time here is about 7 days; lead time is about 26 days. The boat is being worked on a quarter of the time it is in the yard. The three wait boxes — parts, paint bay, electrician — are where a lead wins or loses.
PCE = touch time ÷ lead time. In the map above, 7 ÷ 26 ≈ 27%. Many jobs run under 10%. You do not chase the 27% of touch time — you attack the 73% of waiting. That is where the weeks are.
Value stream mapping, touch time vs lead time, and process cycle efficiency are standard Lean tools (Learning to See, Rother & Shook; Lean Enterprise Institute).
Chapter 3
Name the waits. Each one is a target.
Once the map shows the waits, name them. At ESI the same ones repeat on nearly every job:
| The wait | What it really is |
|---|---|
| Waiting on parts | Long-lead item (gearbox, shaft, seal) or a small consumable (anode, cutless, clamp) ordered too late |
| Waiting on approval | Estimate/change-order sitting for a customer or manager sign-off |
| Waiting on the next trade | The boat is done with one tech and queued for the electrician, painter, or rigger |
| Waiting on a resource | The haul-out slot, the lift, the paint bay — a shared constraint (Chapter 5) |
| Waiting on rework | Something failed the check and loops back — the most expensive wait |
A boat on the hard is occupying a stand, a slot, and the customer’s patience while earning nothing. It is capital sitting still. You do not need an exact number to lead — you need the crew to feel that a $12 anode holding a $40,000 repower for a week is a catastrophe, not a footnote.
Waiting is one of the eight wastes (Tech 1, DOWNTIME); this chapter applies it at the whole-job level.
Chapter 4
Starting more boats feels productive. It makes every boat slower.
Work in process (WIP) is the number of boats torn apart at once. The instinct — and the customer pressure — is to start every boat so everyone sees progress. It is the most common flow mistake a yard makes, and it backfires.
Delivery time rises with the number of jobs in progress. Roughly: lead time = WIP ÷ throughput. Double the open boats without adding capacity and you roughly double how long each one takes. Fewer boats, moving fast, beats many boats crawling.
Little’s Law and WIP limits are core to Lean and Kanban flow management (Reinertsen, Anderson).
Chapter 5
Every flow has one thing that gates it. Manage that, ignore the rest.
In any flow, one resource gates the throughput of the whole thing — the constraint (bottleneck). At ESI it is usually the lift/haul-out slot, the paint bay, or the one person who does a specialised task (the rigger, the alignment tech). Everything upstream can only flow as fast as the constraint lets it.
An hour lost at the bottleneck is lost for the whole yard, forever. An hour saved at a non-bottleneck is a mirage. Making a fast station faster just piles more half-done boats in front of the constraint. Leads waste enormous effort optimising stations that were never the problem.
Theory of Constraints (Goldratt, The Goal): identify, exploit, subordinate, elevate the constraint.
Chapter 6
Never let a $12 part hold a $40k boat.
Chapter 3 named it: waiting on parts is the wait that shows up on almost every job. It is also the most fixable. The fix is pull — make parts arrive ahead of the work, not in response to it.
The big items (gearbox, shaft, GLIDE seal & cutless, prop, paint) have real lead times. They must be ordered the day the job is committed, not the day the tech reaches for them. This is exactly the Repower course’s ‘commit early, pre-order long-lead parts’ point — a lead owns making it happen every time.
The small stuff that stalls jobs — anodes, cutless bearings, sealant, hose, 316 clamps, fasteners — runs on a two-bin (min/max) kanban:
Pull systems and two-bin kanban are foundational Lean/JIT (Tech 1’s JIT pillar, applied to the parts shelf).
Chapter 7
One boat is a task. A yard full of boats is a flow to conduct.
A lead does not schedule one boat — they conduct the whole yard. The goal is a steady flow of finished boats, not a yard full of started ones.
Sequence the work so the constraint (Chapter 5) is never idle and never buried. If the paint bay is the gate, stagger jobs so boats arrive at paint in a smooth stream — not three at once (a jam) and then none (starved).
Chapter 8
Go and see the flow. Don’t manage it from a desk.
Gemba means ‘the real place’ — where the work happens. A Gemba walk is the lead going to the boats, every day, to see the flow with their own eyes. It is not an inspection and it is not you grabbing a wrench — it is looking, asking, and coaching.
You are not checking whether each tech is busy. You are checking whether each boat is moving: What is it waiting on? What’s its constraint today? Is anything stuck? The famous catch is the boat that has been ‘90% done’ for two weeks — busy all around it, but not moving. A daily walk finds it on day two, not week two.
Gemba walks and ‘go and see’ (genchi genbutsu) are core Toyota/Lean leadership practice.
Chapter 9
When flow breaks, don’t band-aid it. Find the real cause.
When the same wait keeps happening — the anode back-order, the alignment that fails and loops back — a lead does not just push through it again. They find the root cause and remove it, so it stops recurring. Two tools do this.
Ask why about five times, each answer feeding the next, until you reach a cause you can actually fix:
The band-aid was ‘expedite the anode.’ The root-cause fix is ‘two-bin kanban on anodes’ (Chapter 6) — and the wait never comes back.
An A3 is a single page (originally A3-size paper) that structures a real improvement so it’s thought through, not guessed:
| A3 section | What goes in it |
|---|---|
| Background | The problem and why it matters (the wait, its cost) |
| Current state | What actually happens now — with the map/numbers |
| Target | The measurable goal (cut this wait from 10 days to 1) |
| Root cause | The 5-Whys result |
| Countermeasures | The change that removes the cause |
| Plan | Who does what, by when |
| Follow-up | Did it work? The after-number (Chapter 10) |
5 Whys (Ohno) and A3 problem-solving (Toyota; Managing to Learn, Shook) are the Lean root-cause standard.
Chapter 10
‘We think it’s better’ is not an answer. The number is.
Tech 1 had no numbers — it was awareness. A lead runs on measurement: you cannot tell whether an improvement worked, or defend it, without a before-and-after number.
| Metric | What it tells the lead |
|---|---|
| Lead time (days in yard) | The headline — how long a boat actually takes to deliver |
| Touch time / PCE | How much of that was real work vs waiting (Chapter 2) |
| WIP (boats open) | The lever from Chapter 4 — watch it, cap it |
| On-time delivery | Did we hit the promise? The customer’s metric |
| Rework / PDI defects | Quality — each one is an expensive loop back |
Before you change anything, write down the current number (the current-state map is your baseline). Make the change. Measure again. If the anode kanban was supposed to cut the parts wait, the parts-wait number must actually drop — or the countermeasure failed and you try again.
Put it where the crew works: a board with every boat, its stage, its constraint, its parts status, and the WIP cap. Red/amber/green a boat that’s stalled. The board makes flow visible to everyone, drives the daily Gemba stop, and turns ‘how are we doing?’ into a glance.
Chapter 11
Most improvements fade in a month. A lead makes them permanent.
The hardest part of Kaizen is not the first improvement — it is that improvements decay. The hero fixes the flow, moves on, and in a month the yard is back to how it was. A lead’s real job is making the gain stick.
An improvement that lives only in one person’s head is temporary. Write it into standard work — the agreed best way, that everyone follows and anyone can improve. The anode kanban, the ‘order long-lead parts at commit’ rule, the WIP cap: each becomes standard, not a favour. Standard work is the ratchet that stops you sliding back.
Keep improving in a disciplined loop rather than random bursts:
Small, steady, measured steps — the Tech 1 Kaizen spirit, now run deliberately by the lead.
None of this survives in a blame culture. If surfacing a problem gets a tech chewed out, they will hide problems — and a lead who can’t see problems can’t fix flow. Respect for people (Tech 1’s second pillar) is what keeps the ** notes coming:
Chapter 12
One boat, one yard, start to finish — the whole method in order.
Pulled together, the lead’s job is a repeatable routine, not a set of tricks:
| Step | The move | Chapter |
|---|---|---|
| 1 | Map each job — work vs waiting, find the biggest wait | 2–3 |
| 2 | Cap WIP — finish before you start | 4 |
| 3 | Find the constraint — keep it fed, subordinate the rest | 5 |
| 4 | Pull parts — long-lead at commit, consumables on kanban | 6 |
| 5 | Schedule the yard around the constraint and the WIP cap | 7 |
| 6 | Walk the Gemba daily — ask what each boat waits on, coach | 8 |
| 7 | Root-cause the recurring stalls — 5 Whys, A3 | 9 |
| 8 | Measure — baseline, after, on the board | 10 |
| 9 | Sustain — standard work, Kata, respect for people | 11 |
Commit the job → order the gearbox/shaft/seal/paint same day (step 4). Don’t open it until a WIP slot frees (2). Map it; the paint bay is the constraint (3,5). Stagger it so it hits paint clean (7). Daily walk: on day 6 it’s waiting on the electrician — named, owned, chased (8). The alignment fails once and loops — 5-Whys it, the root cause is a worn dial gauge, replace it and standardise the check (9,11). Delivered in 18 days instead of 26; lead time logged, kanban that saved the anode wait made standard (10,11).
A lead owns the flow of their section; escalate to the manager (with data, an A3, and a ** flag) when a fix needs money or capacity (a second painter, a new lift day), when a constraint is structural, or when a customer/commercial decision is required. Bring the number and the recommendation, not just the problem.
For leads
The flow traps a yard falls into again and again — and how a lead counters each one.
| The trap | Why it happens | The lead’s counter |
|---|---|---|
| “Start it so the customer sees progress” | Pressure to look busy — starting is visible, finishing isn’t | Cut WIP; finish boats before opening new ones — progress is delivery, not disassembly |
| Optimising a fast station while the bottleneck sits idle | Everyone stays ‘busy’ locally | Subordinate to the constraint; keep the bottleneck fed, let non-bottlenecks wait |
| A $12 anode holds a $40k repower for a week | Long-lead & consumable parts ordered at the step, not at kickoff | Pull: long-lead at commit; two-bin kanban on consumables |
| A boat ‘90% done’ for two weeks | No one owns the last 10%; waiting hidden as progress | Daily Gemba walk of the flow; track lead time, not activity; A3 the stall |
| The same stall every job | Band-aid/expedite instead of root cause | 5 Whys / A3 to the real cause, then error-proof and standardise it |
| The improvement faded in a month | Hero fix, never locked into standard work | Standardise the gain, run the Kata loop, measure it stays |
| ‘We think it’s better’ | No baseline, no after-number | Baseline before, measure after — put it on the board |
| Techs stop raising problems | Blame culture — flags get punished | Respect for people: make it safe, thank the flag, attack the process not the person |
Final certification
A short check ran at the end of each chapter. Pass this final at 80% to unlock your certificate.