Implementation Guide
The 7 Wastes of Lean.
The 7 wastes of Lean, remembered as TIMWOOD or DOTWIMP, are defects, overproduction, transport, waiting, excess inventory, motion and overprocessing: the seven ways a process consumes time and money without adding value. They are a seeing vocabulary for diagnosis, not a to-do list. And one of them outranks the rest: inventory buys silence for all the others, which is why reducing it feels dangerous, and why it works.
What the 7 wastes are
Lean defines value with three tests, all taken from the customer's side: the customer would pay for the step, the step physically changes the product toward what was ordered, and it is done right the first time. Everything else that consumes time, material, movement or attention is waste, muda in the Toyota vocabulary. Taiichi Ohno sorted that waste into seven recurring categories, remembered today by the mnemonic TIMWOOD: transport, inventory, motion, waiting, overproduction, overprocessing and defects. The same list is also taught as DOTWIMP, with defects first and inventory spelled out as excess inventory.
The list is not a compliance framework and not an improvement program. It is a vocabulary for seeing. Untrained eyes walk past waste every day because waste dresses as work: a forklift crossing the hall looks busy, a full pallet rack looks like security, a machine running looks like productivity. Once a team can name the seven categories, the same walk through the same plant shows a different building.
Muda also has two siblings that are mentioned less and matter more: mura, unevenness, and muri, overburden. Much of the muda you will find is downstream of them. A schedule that swings hard from week to week forces buffers, overtime and waiting into existence; a machine or a person loaded past reasonable capacity produces breakdowns and defects. Hunting waste piece by piece while the schedule stays lumpy is mopping the floor with the tap still running. This page focuses on the seven, but a diagnosis that ignores mura and muri will misread half of what it sees.
Where the lens sits in the roadmap
On the TeamGuru deployment roadmap the seven wastes belong to current-state assessment, at the very start of the Diagnose stage. They are the seeing vocabulary the whole stage is built on: gemba walks use them as themes, the assessment uses them to structure its loss analysis, and value stream mapping converts the biggest ones into a measured picture of flow. Learn to see first; measure second; choose priorities third.
- You are here Current-State Assessment
- In parallel Value Stream Mapping
- In parallel Baseline KPIs
- After Strategy Deployment
Why inventory is the master waste
The classic teaching, and it is correct, is that overproduction is the worst of the seven because it manufactures the other six. Produce more than the next process needs, or produce it earlier, and the surplus must be transported, stored, counted and financed, walked around, and searched through. And when a defect is finally discovered inside the pile, it is discovered weeks after it was made, with the trail gone cold.
Overproduction, however, is hard to catch in the act, because it looks virtuous while it happens. A running machine is earning its keep; an operator ahead of schedule is a good operator. What overproduction leaves behind is countable: inventory. Inventory is overproduction's visible twin, the frozen record of every pushed schedule and every oversized batch, and that is what makes it the practical entry point of a diagnosis.
Inventory also plays a deeper role: it buys silence for all the other wastes. With six days of blanks in front of machining, a breakdown costs nothing today. A 47-minute changeover disturbs nobody. A bad batch from the upstream process is discovered six days late, when the evidence is cold and the argument is unwinnable. Every pile between two processes is an insurance premium, paid daily in cash, floor space and lead time, against problems nobody is then forced to fix.
That is why reducing inventory feels dangerous: it genuinely withdraws the insurance, and the hidden problems surface fast. It is also exactly why it works. Lower the level deliberately, one buffer at a time, with problem-solving capacity standing by, and the plant will tell you what is really wrong faster than any audit. Feeling exposed is not a side effect of the method. It is the method.
The seven wastes, with examples
The table walks all seven in the DOTWIMP order, plus the commonly added eighth, through the illustrative 450-person components manufacturer used across these guides: 456 units a day through cutting, machining, welding and assembly, two shifts, 54,000 seconds of daily working time, a takt of 118 seconds. The numbers are illustrative but internally consistent with the value stream mapping guide, which maps the same plant.
| Waste | What it looks like at the plant | How to spot it on a walk | Counter-practice |
|---|---|---|---|
| Defects | Leaking fittings caught at final test loop back to assembly for rework, consuming a second dose of takt time per failed unit and a permanently staffed rework bench. | Find the red bins, the rework benches and the second inspection stations. Count what is in them; every piece is the process talking. | Error-proof the escape points and run root cause work on the repeaters instead of inspecting harder. Poka-yoke guide RCCA guide |
| Overproduction | Cutting produces to the weekly MRP schedule whether machining consumed its last batch or not. The efficient-looking extra output is exactly what fills the 6-day pool. | Look for a process running hard while its downstream queue is already full. That is overproduction happening live, not a symptom of it. | Pull systems designed on the value stream map, so nothing is produced before it is consumed downstream. Value stream mapping guide |
| Transport | Cut blanks ride a forklift 260 meters from the saw to machining, wait in a staging lane, and are handled three more times before a spindle ever touches them. | Follow one pallet from process to process. Count the moves, the meters, and every time it is set down and picked up again. | Bring sequential processes closer together, and give every item one addressed home so moves stop being search trips. 5S guide |
| Waiting | Welding runs at roughly half of takt, so the welder finishes in about a minute and waits out the rest of every 118-second cycle. Assembly, meanwhile, waits each Monday for the weekly schedule release. | Stand at one station for ten cycles and note every wait with its reason: parts, a crane, an inspection, a decision. | Balance the line against takt and standardize the sequence so waits become visible exceptions. Standardized work guide |
| Excess inventory | A 6-day pool of blanks, roughly 2,700 pieces, sits in front of machining because the 47-minute changeover forces batches of about three days of demand. | Count the pieces between any two processes and divide by daily demand. At 456 units a day, 2,700 pieces is six days of lead time doing nothing. | Shrink the batches that create it, usually via changeover reduction, then cap what must remain with supermarkets. SMED guide |
| Motion | An assembly operator walks 14 meters per 118-second cycle to fetch fittings from a shared shelf, about 3 kilometers over a shift that adds nothing to the product. | Watch ten cycles at one station and split the time: hands on the product versus reaching, walking and searching. | Point-of-use material presentation and a workstation organized so the normal reach is the short one. 5S guide |
| Overprocessing | Every housing gets a full dimensional check at machining although the drawing requires it only for one safety-critical variant, and welds are ground smooth on faces no customer ever sees. | Pick one step and ask who asked for this. If the answer is not the customer, a regulation, or the next process, it is a candidate. | A standard that defines the required level of processing and checking, no less and no more, agreed with quality. Standardized work guide |
| Unused talent | The operator who has reset the same sensor fault three times a shift for a year, and has never once been asked what causes it. | Ask three operators what they would fix first. If the answers come fast and specific, the waste is not on the floor. It is in the channel. | A working improvement system that captures ideas at the point of work and implements most of them. Kaizen guide |
The eighth row is a later addition to the canon, and it earns its place. A plant that never asks its operators what they see is running its cheapest and best-placed sensors disconnected.
How to run a waste-spotting walk
A waste walk is a themed gemba walk: 30 minutes, at the process, and it produces numbers and quotes, not impressions. The sheet below gives ten observable prompts. Do not attempt all ten in one walk: pick three or four, walk with a partner, and agree on what you both saw before writing anything down.
| # | Prompt | What it exposes |
|---|---|---|
| Count the pieces sitting between two stations, then divide by daily demand to convert the pile into hours or days. | Excess inventory | |
| Time one operator for ten cycles: hands on the product versus walking, reaching and searching. | Motion | |
| Find the oldest date in the WIP area, on a label, a traveler or the dust, and ask why that part is still here. | Excess inventory, often defects | |
| Stand at one station for ten full cycles and note every wait with its reason: parts, crane, inspection, decision, information. | Waiting | |
| Follow one pallet from a process to the next one. Count the moves, the meters and the set-downs. | Transport | |
| Find a process running while its downstream queue is full, and ask what the operator has been told to do when the queue fills up. | Overproduction | |
| Locate the rework bench and the red bins. Count the contents and read the defect tags. | Defects | |
| Pick one processing or inspection step and ask who requires it: the customer, a regulation, the next process, or habit. | Overprocessing | |
| Watch one changeover end to end, or ask when the last one happened and how long it took. | The batch logic behind inventory | |
| Ask one operator what they would change first, and write the answer down word for word. | Unused talent |
Three rules that keep the walk honest
- Wastes are symptoms, so trace each one to its system cause before proposing anything: transport points at layout, inventory at batch logic and changeovers, waiting at line imbalance, defects at process control.
- Study the process, never the person. The operator walking three kilometers a shift did not design the workstation, and the cutting crew building the pile is following the schedule it was given.
- Do not launch seven crusades. The walk feeds the diagnosis; the diagnosis, not the walk, decides where improvement effort goes.
On the roadmap, all of this lands in the Diagnose stage: the walk observations, together with measured KPIs and the value stream map, become one shared picture of where the plant loses time and money.
From spotting to acting
An observation becomes useful the moment it gets a rough number. The 2,700 pieces in front of machining are six days of lead time and six days of cash. The 14 meters of walking per cycle are roughly half an hour of every operator shift. A rework bench holding thirty units is a day of defects made visible in one place. Quantify in hours, pieces and meters, not in philosophy, and accept rough: a count taken at the process today beats a precise report about last month.
During Diagnose, those quantities aggregate into the assessment's loss analysis: the two or three biggest loss buckets, roughly costed, become the case for change and the shortlist that value stream mapping and the KPI baseline dig into. Later, once a daily management rhythm exists, the same lens becomes a permanent source of improvement ideas, because everyone who can name the wastes spots them at their own station.
Either way, findings need a home, because a walk observation that dies in a notebook teaches people to stop mentioning what they see. In TeamGuru, walk findings become owned, dated actions and improvement ideas enter one kaizen pipeline with visible status, so the walk leaves a trail instead of a memory.
Common mistakes
What bad looks like
- Waste safari tourism: a day of photos, a slide deck, no owner, no follow-up
- Operators blamed for waste the system forces on them: the walking, the waiting, the batches
- Overprocessing hunts that cut real quality steps because a check looked redundant from the office
- Seven simultaneous crusades launched from one enthusiastic walk
- Waste percentages announced from ERP reports nobody verified at the process
What good looks like
- Themed walks: one waste, 30 minutes, at the process, with a partner
- Findings quantified the same day in pieces, hours and meters
- Every observation traced to a system cause before any countermeasure
- The two or three biggest losses feed the assessment and the value stream map
- Operators asked what gets in their way, and thanked for the answer
What comes next
The seven wastes are the beginning of a diagnosis, not the whole of one. What the walks surface flows into two harder-edged instruments: value stream mapping, which turns the biggest wastes into a measured picture of flow with a future state and an improvement plan, and the baseline KPIs that make the losses trend visibly, week after week, instead of living in one workshop's memory.
From there, the roadmap continues into strategy deployment, where the largest quantified losses compete for a place among the plant's few real priorities. A team that has learned to see its waste, counted it, and cannot yet act on all of it has done the stage correctly. The seeing comes first; the system that acts on it is what the rest of the roadmap builds.
