Free Trial

← Back to blog

Production managementReading time 11 min

The seven wastes, and how to actually spot them in your own plant

Waste is hard to remove because most of it looks like work.

7

01The seven wastes

WasteWhat it looks like in a shopHow to measure it
TransportA semi-finished part crosses the plant three times to reach the next stationTravel distance, number of moves
InventoryWIP stacked in the aisle, finished goods sitting two yearsWIP value, age distribution
MotionThe operator walks to the far end for a gauge, every single timeOperator walking distance, retrieval time
WaitingWaiting on drawings, material, inspection, a supervisor's signatureWaiting hours as a share of total
OverproductionFifty extra pieces made to round out a batchQuantity released ÷ quantity ordered
OverprocessingPolishing a non-functional face to Ra 0.8Compare against what the drawing asks for
Defects and reworkReworking a dimension that went out of toleranceDefect rate, rework hours
An eighth is often added later: unused employee talent.

02How to see it

Value stream mapping

Map one representative product's full path from order to shipment, marking process time and wait time at every station, then compute the value-added ratio = process time ÷ total lead time.

  • In most shops that have not been through improvement work, that ratio lands between 0.5% and 5%.
  • Which is to say: 99% of the time, the part is waiting.
  • The room for improvement is almost entirely in the waiting, not in the cutting speed.

Spaghetti diagram

Take the plant floor plan and trace one product's, or one operator's, movement over a day. Lines tangled into a knot mean a layout problem. It is the most intuitive and cheapest diagnostic there is.

Genchi genbutsu

Stand on the floor and watch one complete work cycle. Do not read a report. Toyota's version is to draw a circle and stand in it for thirty minutes, the waiting and the repeated motions never show up on a report, and only appear if you are standing there.

03The matching tools

ProblemToolEffect
Long changeovers force big batchesSMEDSetup time often drops by more than half
Hunting for tools takes forever5SMotion and waiting waste
Too much WIPPull production / kanbanInventory and lead time
Defects escape to the next stationJidoka, mistake-proofingDefects and rework
Layout drives transportCell productionTransport and waiting
The same problem keeps coming back5 Why、A3Root cause never addressed
SMED is usually the highest-return place to start, it solves batch size and inventory at the same time.

The core idea of SMED is to split changeover into internal work (the machine must be stopped) and external work (can be done while it runs), move everything that can be externalized out first, then simplify what internal work remains. The first step alone typically cuts the time in half.

04Common traps

  • Chasing machine utilization. Loading a non-bottleneck to the ceiling only manufactures WIP. See production scheduling.
  • Doing 5S for the photographs. The place looks immaculate, the process never changed, and three months later it is back to how it was.
  • Rolling it out everywhere at once. Pick one line as a model, get a result, then spread it, the success rate beats a plant-wide launch by a wide margin.
  • Treating improvement as cost reduction. The goal is shorter lead time. Cost falling out of it is a result, not the method.