Art of Lean
Back to Reference
Waste Elimination

Waiting

One of the 7 Wastes — idle time when people stand watching machines, wait for parts from upstream, wait for equipment repair, or wait for information. The Japanese term temachi literally means "hands waiting," emphasizing that capable workers are standing with nothing to do.

Japanese

手待ちのムダ

temachi no muda

waste of idle hands; waste of waiting with hands free

Also known as

Waste of Waiting, Idle Time

Definition

Waiting is idle time — when an operator, a machine, or a process is not performing value-adding work because it is waiting for something: parts from an upstream process, completion of a machine cycle, a changeover, an equipment repair, information, or a decision. The waste is not the worker’s fault — it is the system’s failure to keep work flowing to the worker continuously.

At Toyota, waiting is considered a management problem, not a worker discipline problem. If a worker is idle, the question is not “why isn’t that person busy?” but “what is wrong with the production system that it cannot keep work flowing?”

Japanese Origin

手待ち (temachi) is a compound of 手 (te, “hand”) and 待ち (machi, “waiting,” from 待つ, matsu, “to wait”). Literally: “hands waiting.” The image is vivid — a worker standing with idle hands, ready and willing to work but with nothing to do. The term places the emphasis on the human cost: capable people with nothing productive to occupy them.

This is distinct from the concept of leisure or break time. Temachi specifically describes involuntary idleness within the work period — the worker is at their station, during their shift, but the system has failed to provide them with work.

History at Toyota

Waiting has roots in the industrial engineering process chart tradition, where the D (Delay) symbol represented unplanned stagnation — a point where material or an operator sits idle. But the IE tradition treated delay as a descriptive category for process analysis, not as a systemic failure to be eliminated. Ohno reframed it: waiting is not a natural feature of production but evidence that the production system has failed to keep work flowing to the worker.

Ohno’s early experiments at the Honsha Plant in the late 1940s and 1950s directly confronted waiting waste. The prevailing factory model — one operator per machine — meant that operators spent most of their time watching machine cycles. A machine might cycle for 30 seconds while the operator’s manual work required only 5 seconds. The remaining 25 seconds were pure waiting.

Multiprocess handling — Ohno’s countermeasure was to assign one operator to multiple machines. Instead of watching one machine cycle, the operator would load a part into machine A, walk to machine B and unload/load it, then to machine C, and return to machine A as its cycle completed. This required jidoka (machines that could detect abnormalities and stop themselves) so the operator did not need to watch each machine.

By 1954, Toyota had achieved single operators running as many as 17 machines in the machining shops — a dramatic reduction in waiting waste that also transformed labor productivity.

Leveled production — Waiting also occurs when upstream processes deliver parts in uneven batches. If the stamping department produces a full day’s worth of parts in one run and delivers them all at once, downstream assembly has nothing to work on until the batch arrives, then is flooded. Heijunka (production leveling) smooths this by producing in small, mixed quantities throughout the day.

How to Recognize Waiting

  • Operators standing idle while machines cycle
  • Workers at one process waiting for parts from the previous process
  • Machines sitting idle because work has not arrived
  • Entire lines stopped waiting for a changeover to complete
  • Operators waiting for a team leader, supervisor, or quality inspector
  • Processes waiting for information (drawings, specifications, approvals)
  • End-of-shift idle time when work runs out before the shift ends

Countermeasures

Multiprocess handling — Assign operators to multiple machines or processes, eliminating the time spent watching a single machine cycle. This requires jidoka (automatic stop on abnormality) and a U-shaped or sequential machine layout.

Heijunka (production leveling) — Level the production schedule to deliver parts in a continuous stream rather than large batches. This prevents feast-or-famine cycles where downstream processes alternate between waiting and being overwhelmed.

SMED (quick changeover) — Reduce changeover time so that machines are available for production sooner. Long changeovers create waiting for both the machine being changed over and the downstream processes starved of parts.

Preventive maintenance (TPM) — Reduce unplanned equipment breakdowns that cause downstream processes to wait for parts. Autonomous maintenance — operators performing daily cleaning, inspection, and lubrication — catches problems before they become breakdowns.

Standard work and line balancing — Distribute work elements evenly across operators so that no operator finishes significantly before or after others. If one operator’s cycle time is 40 seconds and another’s is 80 seconds, the faster operator waits 40 seconds each cycle.

Common Mistakes

Keeping operators busy with non-value-adding work to avoid visible waiting. Some managers, uncomfortable seeing idle workers, assign busy-work: cleaning that does not need to be done, sorting that adds no value, moving parts that do not need to be moved. This converts visible waiting waste into hidden motion waste or over-processing waste. The underlying system problem remains unsolved but is now invisible.

Confusing waiting with the operator watching the machine. Watching a machine cycle is waiting, even if the operator appears attentive. If the machine has jidoka (automatic stop on abnormality), the operator’s presence during the cycle adds no value. Toyota’s approach is to eliminate the need for watching by building detection capability into the machine.

Ignoring waiting because utilization metrics look good. Machine utilization and labor utilization metrics can mask enormous waiting waste. A machine that runs 95% of the time may be running the right parts only 60% of the time, with the other 35% producing ahead of demand (overproduction). Utilization is not the same as productive work.

Adding inventory buffers instead of solving the root cause. When a process frequently waits for parts, the common response is to add a buffer inventory between processes. This hides the waiting problem but creates inventory waste, transport waste, and capital consumption. Toyota’s approach is to make waiting visible, investigate the root cause (equipment reliability, changeover time, scheduling unevenness), and solve it.