Lean and process improvement
The improvement methods a plant uses to shorten lead time and take cost out of the process, and what each one actually demands of the people running it.
- 5S: the disciplines underneath the photographs
What 5S asks of operators and supervisors, why sorting is the argument, and how to distinguish a sustained standard from a cleaning campaign.
- Andon: the escalation promise behind the light
What an andon call commits management to, why call rates collapse to nothing, and how to set response tiers that survive a busy shift.
- Autonomous maintenance: when the operator owns the machine's condition
Cleaning as inspection, restoring basic condition, the abnormality tag queue, and the competence boundary deciding what an operator may touch.
- Defects as waste: counting the full bill for work done twice
Why scrap value understates what a defect costs, how the point of detection changes the bill, and how rework quietly disappears from the figures.
- Flow before pull: why the order of these two principles decides the outcome
Why flow is a claim about queues rather than speed, why pull over an unstable process backfires, and how to test whether either is applied.
- Gemba walks: looking at the work without turning it into an inspection
How to walk the floor so people tell you the truth: a route and rhythm, questions instead of instructions, and a ledger for what you found.
- Hoshin planning: choosing the few objectives the plant will genuinely pursue
Policy deployment as an exercise in subtraction: negotiating targets down the line, separating breakthrough work from daily management, reviewing honestly.
- Improvement in a jobbing shop: what repeats when the products do not
In low-volume work the process repeats even when the part does not. Which methods transfer, which do not, and where the real losses actually sit.
- Improvement in process plants: the material already flows, so where is the waste?
In continuous and batch processing the losses are yield, giveaway, off-spec and cleaning time rather than queues, and campaign length is the batch argument.
- Improvement kata: a practice routine for reaching a condition you cannot yet see
Target condition, obstacles and single experiments, plus what the coaching routine costs a manager and why it is hardest for good problem solvers.
- Improving flow: finding where work stops and deciding what to attack first
Work spends most of its life waiting. How to follow one product family, rank the stoppages, and avoid changes that only relocate the queue.
- Kaizen events: the concentrated week and the preparation it depends on
Scoping a focused improvement workshop, releasing the right people, preparing parts and permissions, and knowing when an event is wrong.
- Kaizen: small improvements, and the management response time they depend on
The daily habit of small operator-led change, what supervisors owe in return, and why unanswered ideas end the practice faster than any failed trial.
- Lead-time reduction: shortening the clock the customer actually experiences
Order-to-despatch time is mostly queue, engineering and procurement. How to measure it honestly, and why quoted lead times protect themselves.
- Lean implementation: what has to change in how the plant is managed
Sequencing a lean start, choosing a first area, the management habits that must change, and why tool-led rollouts stall once the coordinator leaves.
- Mistake-proofing: designing the error out instead of asking for more care
Why prevention beats detection, how mistake-proofing devices get bypassed, and how to confirm the device still catches what it was built for.
- One-piece flow: removing the queue between operations and living with what that exposes
What single-unit flow demands of reliability, balance and changeover, the utilisation you trade away, and where a batch of one is physically impossible.
- Over-processing: effort the customer never asked anyone to spend
Tolerances tightened by habit, checks added after an incident and never removed, and why nobody wants to be the person who deletes a process step.
- Overall equipment effectiveness: what the measure is for, and how it gets gamed
Why the three loss factors multiply, why the figure is only comparable with itself, how definitions get manipulated, and when chasing it destroys value.
- Overproduction: the waste that looks like a good day's output
Why making early is as costly as making too much, the measures that reward it, and what surfaces the moment a plant stops building ahead of demand.
- Production kanban: designing a signal the floor will actually obey
Choosing the carrier for a production signal, setting loop quantities without false precision, handling exceptions, and keeping cards honest.
- Production levelling: deciding where demand variation is going to be absorbed
Levelling moves demand variation into stock, backlog or lead time instead of the shop floor, and presents a changeover bill the plant must be able to pay.
- Proving an improvement was real: baselines, freed time and savings that never arrive
Telling genuine improvement from a tidier floor: fixing the baseline first, treating freed time honestly, and not counting activity as a result.
- Pull systems: letting consumption authorise production instead of a forecast
How pull moves the release decision downstream, which architecture suits which mix, what it demands before it works, and where pull is the wrong answer.
- Quick changeover method: dividing the work that needs the machine stopped
Recording a changeover before arguing about it, splitting internal from external work, converting and streamlining, and why gains decay.
- Rearranging an existing plant for flow: what it really costs to move a machine
Trialling a rearrangement before lifting anything, the services and safety work deciding what can move, and why a cell is an organisational change.
- Running an improvement programme: pipeline, funding and management attention
Structuring an improvement pipeline, triaging a backlog, funding small changes quickly, and reporting benefits without inventing savings nobody can find.
- Running setup reduction as a programme rather than a one-off event
Choosing which machines to work on, stopping improved changeovers from drifting back, and deciding what to do with the capacity you released.
- Seeing waste: the observation discipline behind the categories
How to watch an operation properly, why the waste categories are a prompt rather than a scoring sheet, and what to do with a list that is far too long.
- Standard work: the current best method, agreed by the people who run it
Fixing sequence, rate and in-process stock, writing it from observation rather than intention, and auditing without policing operators.
- Stock on the shop floor as a symptom of something else
Reading each pile of work as evidence of a specific problem, lowering levels deliberately to expose it, and resisting the urge to add storage.
- The single-page problem report: forcing a thinking sequence onto one sheet
Why the whole problem has to fit one sheet, how these reports get written backwards from a chosen solution, and what the follow-up half is for.
- Total productive maintenance: equipment condition as a production responsibility
The bargain inside TPM: what it asks operators to take on, what it asks skilled maintainers to give up, and why pillar rollouts stall.
- Transport and motion: material being moved versus people reaching
Why moving material and moving operators are separate problems with different owners, how to map both cheaply, and which fixes are worth a layout decision.
- Underused skills: the waste that eventually shows up as turnover
What a plant forfeits when only managers are expected to think, the conditions that make participation possible, and why reward schemes distort it.
- Value stream mapping: the walk, the argument and the map nobody reads
Picking a product family, collecting data by walking rather than querying, forcing the honest current state, and turning a future state into owned commitments.
- Visual management: making the state of work obvious without asking anyone
How visual controls show abnormality at a glance, why most boards become decoration, and what has to be defined before a display means anything.
- Waiting: telling idle people apart from idle work
Why material waits far longer than people do, how to find where the waiting happens, and why filling operator idle time with production makes things worse.
- Why improvements come undone, and what actually holds a gain
Reversion is the default. Ownership, confirmation routines, staff turnover, and changing the measure people are judged on before the change decays.
- Work instructions somebody will actually use at the machine
Writing for the reader rather than the auditor, marking the critical steps, handling language and shift reality, and controlling revisions on a working floor.