Time Study V Actual Production 2

[ProcessDave]

When we complete a time study, the production from the floor never achieves the theoretical volume the time study indicates we should get. When an operation is not controlled by another process such as a machine opening and closing, operators doing "Bench Work" achieve a percentage, say 85% of what a time study indicates. Is there an industry accepted standard for this, and where could I find information on it?

 

[BML]

Generally speaking, a process will never achieve 100% efficiency because there is always something that goes "wrong". Perhaps this could be an occasional shortage of parts, or an unfocused operator, or even bathroom or meeting or scheduled maintenance breaks. The 85% number has somewhat become the defacto standard. I've personally been able to use it with reasonably good success. One can even show that in many cases, processes that consistently achieve higher than 85% efficiency either (a) have a standard different than what was orignally set due to an incorrect time study or process improvements, or (b) the process is not being followed as originally established. This can be especially true of processes that are operator dependent.

 

[IEFrank]

BML,
Do you find that 85% is the defacto standard after PF&D is accounted for?

 

[BML]

IEFrank, the 85% has been something of a standard for many of the processes I have worked on. One company I worked for began projects with 15% PF&D and 15% inefficiency factors. It was the engineer's job to figure out how to reduce the inefficiency, but not the PF&D.

Again, it also depends on the type of process. Most of my experience has been with operator-dependent assembly or fabrication, so this rule of thumb is much more applicable. In other situations, such as automated circuit board assembly, the 85% rule may not apply as well. Even then, the type of shop (job, high-volume/low-mix) and the robustness of the material handling and communication systems will play a part. A good example is a powdercoat paint line. If you only need to paint one or two colors, then your efficiency should get up to about the 80%-90% level. But if you need to paint 20 colors and 3 different formulations, then you may be lucky to achieve 50% efficiency.

I suppose the ultimate answer to the original question is: No, there is no industry standard for maximum observable efficiency. I would suggest 85% as a starting point. From there, study the process to figure out what contributes to the factor, and adjust as necessary.

 

[IEFrank]

Thank you. It has been so long since someone asked me to set production standards I forgot how to explain the 85%. When I worked in a manufacturing setting, this number was easier to justify and I believe that the time studies were much easier to do. I now work in a distribution center where I believe the time studies are much more difficult to complete because of variablitily in cycles and the 85% efficiency is a more difficult to explain to management (at least for me).

 

[Echeverri]

ProcessDave:

It seems that your time standards have too much variation; this is what is contributing to your efficiency be at around 85%.
§ Your observed times should have less than 3% variation; most of your variation comes from non-value added activities. Eliminate them or at least reduce them to a minimum before you establish a job standard.
§ Also, how accurate are your effort rating and the allowance for incidental activities? These are another sources for error in your measurement. Again, incidental activities are non-vales added, but you need them for your capacity planning; keep these activities separate and well defined for each job. In many instances you may minimize their impact; for instance in an assembly line or sequential job, size the wip (Kanban) between stations of non-bottleneck operations so that those operators can have incidental time, i.e., water and stretch breaks.
§ Manage your bottlenecks well. They keep the pace of your product or process line. Your manufacturing facility can only make as much product as these ‘pacemakers’ can.

 

[ietech]

Just my $.02

I use the 85% rule as a basis for comparison, I have had studies that exceeded and some below that were good representations.

You probably do all of the following already but I will post anyway just in case there is something here you can use.

I will assume you are including PF&D and proceed from there.

How many samples/observations do you include in your studies?
--- More samples will give a higher accuracy potential.

How many operators do you observe and how many observations per operator?

--- Always exclude the very fast and very slow results. They are either exceptional operators that a normal operator cannot do or possibly an intentional attempt to sabotage your study.

Do you study each major task as a whole or do you break it into its elements.

--- Studying each element of each task is very important and will help you identify possible variables more easily. Again, the more observations the better. Total the results for the elements to get your task bottom line. Total tasks for the completed project.

I firmly believe that extreme accuracy, detail, and good documentation are critical to a good time study.

I have seen several operators verbally bashed and a couple even fired because of sloppy time standards. Please, I'm not implying that yours are sloppy, only stating observations made during a 30 year career.

 

[ietech]

An after thought to my original post.

Try posting in the Industrial Engineering Forum --- You might get some additional input on this subject.

Good Luck

ietech

 

[griffengm]

One thing you need to look at is what determines the cycle time in the process you are studying. Some processes are governed by a nonadjustable machine which the operator feeds and inspects - heavily machine and maintenance dependent. On the other hand are those in which no machines are present and only the operator is active. There is a third class that depends on the tools/setup provided.
A second point is the difference between production time for the shop and time spent on an operation on an individual item. Total time for the shop will likely never achieve 100%, operation time for an individual operator should on occasion surpass it.
A study I came across 4-5 years ago stated that shops with no standards (government) will typically operate at about 40-60% efficiency, shops with historical standards at about 60-70, shops with studied standards but lackluster monitoring at 75-85, and shops with studied and monitored standards at 85-95%.
Hope this helps.

Griffy

 

[flexibility]

ProcessDave (Automotive) Oct 1, 2003
I wonder how are you performing your time study, and for what purpose.
For machine controlled (automatic machine) you cannot assume that the theoretic machine pace will be the production rate. You might include breaking, adjustment and personnal needs. Therefore with automatic machine you might find about 80-85% efficiency, but dont be amazed to find about 110-130% efficiency. The operators have their own tricks how to do it (if they have the motivation).

Operator control operations production might ange from 50 to 130%. The higher value is if you have and incentive plan in the shop .

By definition Time study standard is for average effort but shoud allow an operator to reach 125% efficiency.