Lean Fundamentals
The Seven Wastes of Lean Production: From Shop-floor Identification to Data-supported Improvement
The direct answer for the seven wastes is to classify activities that do not create customer value as overproduction, waiting, transportation, overprocessing, inventory, motion and defects, then use shop-floor observation and process data to locate where they occur, how long they last and what triggers them. Classification alone does not improve anything. Waste moves from being seen to being reduced only when it is connected to ownership, standard work, abnormality handling and a regular review cadence.
Align on the Judgment Standard First: Customer Value, Not How Busy People Look
Whether an activity is waste does not depend on whether employees are busy or equipment is running. It depends on whether the activity changes the product's form or function, meets a customer requirement and is done right the first time. Necessary but non-value-added inspection, changeover or compliance records cannot simply be deleted; pure waste must first be distinguished from auxiliary activities that are necessary under current conditions.
One phenomenon can fall into different categories. Accumulated stock in a warehouse usually appears as inventory waste, but its root cause may be overproduction, distorted planning or waiting for quality decisions. Classification organizes observation rather than establishing one label to argue over. Improvement work must continue to investigate the generating mechanism and constraints.
- First define the observation target: product family, process segment, shift and time window.
- Then record the facts: what happened, how long it lasted and which object was affected.
- Finally state hypotheses: triggering conditions, systemic causes and data needed for validation.
Shop-floor Signals and Data Evidence for the Seven Wastes
The seven wastes are not seven independent reports. Shop-floor identification should examine human motion, material flow, equipment events and quality results together within the same process segment, rather than drawing conclusions from output or inventory alone.
Data validates observation rather than replacing it. MES, SCADA or manual checks can provide timestamps and quantities, but motion waste and overprocessing often require physical walk-throughs, work-element breakdowns and process reviews beyond video or system records.
| Waste type | Typical shop-floor signal | Verifiable evidence | Priority question |
|---|---|---|---|
| Overproduction | Production continues when the downstream process does not need it | Work-order pace, work in process and downstream calls | Who triggered production ahead of need? |
| Waiting | A person, machine or material sits idle waiting | Status events and material-call or maintenance timelines | What condition is being awaited, and who can release it? |
| Transportation | Repeated transfers, long distances or temporary storage | Logistics routes, scan transactions and location changes | Why can it not reach its destination in one move? |
| Overprocessing | Processing beyond customer or process requirements | Process version, inspection specification and work breakdown | Is the source of the requirement still valid? |
| Inventory | Raw material, work in process or finished goods exceed normal buffers | Lot age, location and accumulation reasons | What variation is the buffer concealing? |
| Motion | Searching, bending, repeated confirmation and detours | Motion breakdown, workstation layout and check records | Does the standard workstation support retrieval and placement in one motion? |
| Defects | Rework, scrap, sorting and repeated inspection | Nonconformance codes, repair routes and process parameters | At which step was the defect first created? |
Replace General Patrols with One Structured Shop-floor Observation
Select a product family and process segment with clear boundaries and walk backward from customer or downstream demand. Follow one product, one operator and one abnormality, recording interruptions to flow, human motions and how information triggers the next step respectively.
The observation sheet must allow entries such as ‘unknown’ and ‘to be validated.’ Filling gaps with experience-based causes just to complete the form will only embed guesses in a later digital system. For recurring problems, use equipment events, station transactions or quality data to validate frequency and conditions.
- Prepare: confirm the product family, demand pace, process route and shift-abnormality context.
- Follow the product: record every stop, transfer, split, merge and identity change.
- Follow the person: record retrieval, placement, searching, confirmation, waiting and rework motions.
- Follow the abnormality: reconstruct the timeline from first signal through response, recovery and cause closure.
- Form hypotheses: separate immediately correctable problems from those requiring cross-functional validation.
What Digitalization Should Validate and What It Should Not Replace
Equipment status, scan transactions, Andon events, weighing records and energy curves are suited to building objective timelines and identifying waiting, defects and abnormal transportation. The first value of digitalization is making facts reviewable and using the same definitions before and after improvement, not assigning a sensor to every waste by force.
Overprocessing and motion usually require process and work studies, while inventory and overproduction are also influenced by planning rules and performance mechanisms. Software cannot automatically decide which activities create customer value, nor can it replace cross-functional agreement on standards, ownership and abnormality rules.
| Data object | Question it is suited to answer | Common misinterpretation | Required shop-floor supplement |
|---|---|---|---|
| Equipment status and cycle time | When do waiting and speed losses occur? | Treating every idle period as a loss | Planned, changeover and no-demand states |
| Station and location records | Where do transportation and stagnation occur? | Counting normal process waiting as abnormal | Process dwell and buffer standards |
| Andon and maintenance events | Where does the abnormality-response chain stall? | Evaluating a team only by time to closure | Separate arrival, waiting for parts and production recovery |
| Quality and repair records | Where does a defect first appear and where does it flow? | Counting only the final-inspection detection point | Distinguish the creating operation from the detecting operation |
| Energy and operating-condition curves | Do idling and abnormal consumption accompany waste? | Inferring improvement benefits directly from electricity use | Output, product-mix and environmental baselines |
Move from a Waste List into the Improvement Loop
Improvement priority should consider not only frequency but also customer impact, quality and safety risk, duration, recurrence and implementation constraints. A frequent minor problem and a rare problem affecting critical delivery cannot be ranked mechanically with one rule.
Every improvement should retain its original baseline, target condition, responsible role and validation window. After implementation, check whether waste has moved elsewhere: for example, whether reducing local inventory creates downstream waiting, or whether a shorter changeover causes quality confirmations to be missed.
- Define the problem: describe the object, place, time and impact rather than writing ‘strengthen management.’
- Confirm the mechanism: distinguish direct causes, systemic causes and hypotheses that cannot yet be verified.
- Select countermeasures: eliminate the source first, then reduce occurrence, and only then consider prompts and containment.
- Validate on a small scale: define the comparison basis and observe whether adjacent processes develop new waste.
- Standardize and review: update standard work, abnormality rules and training, and check recurrence on a regular cadence.
Avoid Turning the Seven Wastes into a New Formality
Common failures include teaching classification without changing trigger mechanisms, displaying rankings without allowing teams to explain context, and pursuing complete data while ignoring customer value. In particular, employee busyness should not become a lean objective because it can encourage early production and more inventory.
More robust governance has shop-floor, process, quality, equipment and planning roles jointly review critical problems. Digital dashboards retain definitions and data-quality notes, while improvement records show unresolved issues and applicable boundaries. This makes the seven wastes a shared language rather than a new performance label.
- Do not replace the specific mechanisms of seven problem types with one aggregate metric.
- Do not treat necessary quality confirmation, compliance records or safety steps as waste that can be deleted at will.
- Do not publish simplistic cross-line or cross-plant rankings when data is incomplete.
- Do not present improvement results as unverified benefit commitments.
Bring Classification Back to the Shop Floor and Put Data in Service of Improvement
The seven wastes provide an observation framework: starting from customer value, identify overproduction, waiting, transportation, overprocessing, inventory, motion and defects. Effective work verifies facts along the paths of product, people and abnormalities to find the mechanisms that allow waste to persist.
Digitalization can provide evidence of time, quantity, identity and events, but it cannot replace process judgment and organizational coordination. Start with one process segment to establish a baseline, close the loop through small-scale countermeasure validation, standardization and review, and then decide which data deserves continuous automatic collection.
Frequently asked questions
Should ‘underutilized talent’ be added as an eighth waste?
It can be a useful additional perspective for checking whether employee experience informs improvement and whether recurring problems are judged only by management. This project organizes its content around the classic seven categories; an additional category should not obscure their basic observation methods.
Inspection and changeover do not directly add value, so should both be eliminated?
No. They may be auxiliary activities required by current process, quality or safety conditions. Improvement should confirm requirements, shorten duration, reduce variation or optimize scheduling, not delete steps without considering risk.
Must MES be implemented before the seven wastes can be identified?
No. Shop-floor observation, work-element breakdowns and simple time records are enough to begin. Define the MES, SCADA or Andon data-collection scope only when continuous validation of equipment, lot or abnormality timelines becomes necessary.
In what order should the seven wastes be improved?
There is no fixed order. Prioritize according to customer impact, quality and safety risk, generating mechanism, improvement constraints and effects on adjacent processes. Overproduction and inventory often warrant early examination because they can conceal other problems, but this still requires shop-floor evidence.
