Need advise on creating a time study/work measurement

[casino66]

I'm working on developing a time study/work measurement of the job activities of our warehouse personnel. We need to develop the methodolgy so that we ensure our results are not biased.... Does anyone have a standard methodolgy they could share???

 

[gerhardl]

You do not state the purpose of your study. Work/time studies can be controversial and give negative effects if not properly organized and presented/explained for the organization affected. A lot of literature and studies originated in the periode 1940's to 1970's. After this methods leaned more to human development and group organizing.

Suggestion of checklist:
Purpose and targets to be defined
Timeframe to be decided both for initial check and later rechecks and and development checks
Timeframe and type of for actions to result from checks
One practical methode for the analysis often used is defining all actions to be analyzed (example for warehose: packing of goods, picking items, registering of sending data etc. etc), then registering all people involved in these actions, then establish a mathematical 'non systematic ' (numbers out of hat or existing tabels for this purpose) time list used to visit and check the activities of each single person at that time.

The idea is that this repeated a sufficient amount of times and summed for each activity will give a 'neutral and statistical' picture of the work distribution.

If improvements are the purpose of the analysis the more direct (Scandinavian) approach could give better results in all aspects: ask the employees directly, either to evaluate their time distribution themselves, or to give their three to five suggestions on what single changes they thought would give the best gain to the process and company within their activity area.

 

[tomwalz]

The only way I ever got really good results was to take all the work done and divide it by all the time on the time clock. E.G. packages shipped in a week divided by hours worked in a week. Ask accounting to help. We track dollars generated by dollars spent because that data is readily available. It is more complex but can uncover false assumptions on bids and pricing.

Other wise there was more or less sample prejudice. (Remember the classic GE study with light levels.)

I once worked for a manager who would go out and set standards. He would go out and work like crazy for ten minutes in a perfectly prepared work station, count the results and call that the standard. No one ever met the standard which proved to him how superioe he was over the “hourlies”.

Sometimes you learn much more from bad managers than good ones.


Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessors.com

 

[plasgears]

I found that some operators would take a break when I set up for time study. Or they would automatically go into make-work to throw off the time study.

It's imperative to carry out time study without the knowledge of the subject. a mezzanine would have helped.

 

[jamtilimpact]

I do my time/work studies a lot like tomwalz. I check the database for pieces processed over a three month period and divide it by the hours worked at each operation over that same period. Works like a charm. Occasionally the operators complain that the rate is too high so I regularly do standard stopwatch studies to verify that I gave them a workable rate. They're usually surprised to find that they can actually exceed the rate that I originally set even with allowances. This leads them to start looking at how to improve their work stations and work methods. Better they do that than me.

 

[PSE]

In concurrence with previous posters, check time standards for similar activities in the database (assuming there is one) over a pretty long time line. If you have seasonal or periodic surges or lulls in activity, you will want to make sure they do not adversely skew your data. Do some active "sampling" of the process with the realization that people tend to change habits when they know they are being observed. Get their buy-in, especially when it comes to ideas for improvement.

Regards,

From Murphy's Laws:

To estimate the time it takes to do something, take the amount of time it would take YOU to do it, multiply by 2 and raise it to the next higher unit of measurement.

 

[gth835x]

get a stop watch or two, and make a list of possible things each employee could, should, and shouldnt be doing. Every minute or two, look up, spot check the employees and put a quick check by the category that best fits what their doing.... Go back and do the analysis and if your sample size is sufficient you should have pretty reliable information as towards the productivity of each worker