How do you calculate standard time from a time study?
Short answer
Average the stopwatch readings for each work element, multiply by the operator's performance rating to get normal time, and add the elements. Then multiply by one plus the total allowance for fatigue, personal needs and delays. The result is the standard time per cycle, the time a trained operator working at a normal pace needs.
A time study turns stopwatch readings into a time you can plan with. It separates how long the operator took from how long the task should take at a normal pace.
What are the formulas?
- Mean observed time = the average of the cycle readings for an element.
- Normal time = mean observed time × performance rating, summed over the elements.
- Standard time = normal time × (1 + total allowance).
What does a worked example look like?
A task has two elements.
| Element | Mean observed time | Rating | Normal time |
|---|---|---|---|
| A: pick and position | 42.0 s | 110% | 46.2 s |
| B: fasten and release | 18.5 s | 95% | 17.575 s |
| Cycle | 63.775 s |
With a 15% allowance, standard time = 63.775 × 1.15 = 73.34 seconds per cycle, or 49.1 cycles an hour at one unit per cycle.
How many readings do you need?
Enough that the mean is reliable. For a target accuracy k (such as 5%) and a confidence factor z (1.96 for 95%), the cycles required are ((z × s) ÷ (k × mean))², rounded up, where s is the sample standard deviation of the readings. If element A has s = 3.1 seconds, n = ((1.96 × 3.1) ÷ (0.05 × 42.0))² = 8.4, so 9 readings. An element with fewer readings than its required count needs more observations.
What are the limits?
Ratings are judgments made by the observer, and the allowance is a starting point that has to be set for the job and method studied. The sample-size formula assumes readings that are roughly normal and independent, and nothing here checks that the observed cycles represent normal working conditions.
The time study calculator handles up to eight elements, keeps allowances on their own tab and flags elements with too few readings.