Skip to content
Work Measurement and Capacity | Part 5 of 8

Times Without a Stopwatch: Systems, Comparison, Machine Data

Three other ways to a defensible time, what each is good for, and the rule that every one of them has to carry its source.

Not every time can be measured with a watch. The work may not exist yet, the volume may be too small to justify a study, or the cycle may be driven by a machine. Three other sources produce times you can defend, and each has a rule that keeps it honest.

Systems of predetermined motion times

The principle is simple to state. Human work is described as a sequence of basic motions such as reaching, grasping, moving and positioning. Each motion carries a time that was established once by research, and the analyst builds the time by writing down the motion sequence instead of timing the work. Because the pace is built into the data, no rating is needed.

That makes these systems useful exactly where a stopwatch fails: planning an operation that does not exist yet, comparing two method ideas before a fixture is built, and very short repetitive cycles where reading error would dominate.

Systems of this kind are offered under names such as MTM and by organizations such as REFA. They are licensed methods with their own training, data and forms. Use them through a licensed route and let trained people apply them. Do not rebuild their data yourself, and do not copy their tables or codes into your own documents.

Comparison and estimating

The second source is older and still the most used in practice. A new operation is compared with a measured one that is similar, and the difference in work content is estimated and written down.

Done carelessly, this is how wrong times enter a system and stay for a decade. Done properly, it is legitimate. Name the comparison object, write down what is different and by how much, mark the result as provisional, and set a date by which it will be measured.

In area M2 the twenty-eight variants of the known block family were built this way, each marked provisional with a measurement planned for month six. Twenty-four low volume operations below 200 units a year were documented estimates with a yearly review. Neither group was allowed to look like a measured time in any report.

Machine and process data

The third source needs no judgment at all. Cycle times, program run times and counter data are read from the equipment or the control system.

Two rules apply. Read the program time, not the booked time, because booked time carries waiting and loading with it. And separate the manual share: at the test bench S5, the test cycle of 2.65 minutes runs without the operator, while 0.56 minutes of loading, starting, unloading and reading the result is manual work that carries the allowance.

Getting that separation right is what let area M2 see its real constraint. The test bench time of 3.36 minutes per unit is a machine cycle plus a manual share plus a setup share, and only the manual part behaves like human work.

Your own element library

A fourth source grows out of the first three. Elements that have been measured once can be collected and reused to build times for new combinations of the same work, which is how a plant covers a family of variants without a study for each one.

The library ages just like a time does. Keep it under the same validity rules, or it quietly becomes the oldest data in the plant.

The rule all four share

Every time carries its source in the record: time study, machine data, comparison, estimate or element library. A provisional time is marked as provisional everywhere it appears, including in the reports that planners read.

Of the 148 operations in the M2 product group, 96 carried a measured time at the start. That is 65 percent coverage, and knowing which 65 percent is worth more than having a number in every field.

Newsletter

One Lean idea.
Every two weeks.