A stopwatch and a clipboard. How long the work really takes.
A time study is the lean discipline of putting a stopwatch on the actual work and writing down what you see. The technique is older than lean itself, traceable to Frederick Taylor's industrial engineering work in the early 1900s, and it has aged unevenly. Used badly, time study turns into surveillance and operators learn to game it. Used well, it produces the data without which standard work and line balancing are guesses. The difference is in how the study is framed, who participates, and what is done with the results.
"The number people guess for a task is almost never the number that comes off the clipboard. Measure first, decide later."
A time study breaks a task into work elements. Each element starts and ends at a recognizable physical event, the operator picks up the part, the operator places the part in the fixture, the cycle starts, the part is unloaded. Good elements are short enough to be uniform, usually two to thirty seconds, and bounded by events any observer can identify the same way every time.
With the elements defined, the observer watches the work and records the elapsed time of each element on a worksheet for multiple cycles. Ten cycles is a minimum; twenty or thirty is better. The worksheet captures the time of each element on each cycle, the average, the minimum, the maximum, and notes on any unusual events.
The observer is not a critic. The job is to record what happens, including the variations. If an element takes 12 seconds on cycle 1 and 18 seconds on cycle 7, both numbers go on the sheet, and the variation is part of the data. Variation usually has a cause, a tool reach issue, a fixture seating problem, a material that did not lay right, and surfacing it is part of the study's value.
Time-study data feeds three downstream activities. First, standard work and standardized work cannot be built without it. Second, line balancing across multiple operators or stations needs honest element times. Third, value stream mapping uses element-level data to populate the cycle-time fields in each process box.
Imagine a 30-person assembly shop where the line lead suspects that an underperforming third station is bottlenecking the whole line. Output has been below target for three months and the team has been adding pressure on the station-three operator. Before any more pressure, the lead runs a time study.
She defines fourteen elements across the four stations and watches a full day of production. The data tells a different story. The station-three operator is actually the fastest on the line on most elements. The slowness comes from two elements where the station-two handoff drops parts on the wrong side, forcing the operator to walk an extra eight feet per cycle. The station-three operator has been compensating with speed in the assembly elements but the walking time was uncatchable.
The fix is a small fixture change at station two that drops parts on the correct side. Walking eliminated, the line meets target output the following week. The station-three operator gets an apology rather than a performance plan. That is what a time study at small scale buys: honest data instead of guesses, and a fix that targets the actual constraint instead of the assumed one.
A time study is the data foundation for standard work and standardized work. The element-by-element output feeds into a standard work combination table, which sequences manual, walk, and machine time at takt. When balancing work across multiple operators, the same time-study data is plotted as a yamazumi chart to show how each operator's load compares to cycle time targets.
The questions we hear most about this term.
Long-form guides that pick up where this definition leaves off, written for manufacturers running Arda today.
Same-day setup. No distributor lock-in. Zero stockouts. Top teams double revenue in 9 months.
.svg)
.svg)
%201.svg){kind=logo}
.png)
.svg)
.svg)
