101 Kaizen Templates: Takt Time Calculator



By Jon Miller | Post Date: February 4, 2008 11:41 PM | Comments: 13

A reader pointed out that the takt time for posting the 101 kaizen template has exceeded. It is true that we have not been hitting a consistent 1 template per 3.5 days. To meet our goal for 101 of these by the end of 2008 we will need to recalculate our takt time. Conveniently, we have a takt time calculator as a template for this.

There are two sections to this kaizen template in order to make it easy to show two possible takt time scenarios. There may be an average takt time and a takt time during peak volumes. This information is useful in designing production lines or equipment so that the maximum capacity as well as the daily average can be taken into consideration.
takt%20calculator%201.png
The two sheets in the graphic above are used to illustrate that there are two different shift structures with different numbers of working hours and thus different takt times on these days. For example if there was an "A" shift working four 10-hour days and a "B" shift working three 12-hour days for a 7-day operation the takt could be matched by adjusting average demand.

This takt time calculator will calculate takt time for you under a variety of conditions such as break times, number of working shifts, shift structures and working hours per shift. Fill in the white cells. The blue cells will calculate automatically. The sheet is protected so that you don't accidentally delete the formulas. It is not password protected so feel free to select Tool / Protection / Unprotect Sheet if you would like to modify this template.
takt%20calculator%202.png

Back to the issue of non-performance to takt time on this blog. What is the kaizen action to be taken? Of course root cause analysis and corrective action must follow, but an immediate recovery plan is necessary in the mean time. For illustrative purposes the takt time calculator was modified to show the original plan of one kaizen template every 3.5 days, not adjusted to one every 2.1 days based on the remaining number of kaizen templates (95) the remaining number of blogging months (11) in 2008 and a more realistic number of "available days" per month for blogging (20).
takt%20calculator%203.png
A kaizen template every two blogging days may get to be boring, but a promise is a promise so the challenge will be to meet this goal while keeping it fresh and interesting.

Dear Jon
Thanks for another great template, but I think you can reduce the demand over the remaining 11 months of this year by the 6 templates you already posted. Therefore your calculation of Takt increases from 2.18 to 2.32 days best of luck
Regards
Chris

Poster: Chris Nicholls | Post Date: February 6, 2008 7:33 AM

Oops. Thanks Chris. You are right. 95 left to go, not 101.

Poster: Jon Miller | Post Date: February 6, 2008 9:53 AM

Hi Jon this is a really useful and great template, it's going to be really useful to me, i just have one question/slash observation, Planned downtime cells (C10 and C25) are not referenced in any other cell, so either is information we are not going to need/use (waste) or it should be referenced to Net Working time cells (C11 and C26) please clear this out for me.

thanks.

Poster: Alberto Cárdenas | Post Date: February 17, 2008 3:40 PM

Thanks for the observation Alberto. The template has been corrected and uploaded again. The C10 and C25 cells are indeed relevant now they are referenced as part of the calculation correctly.

That should be a good reminder to everyone to check your their work for errors after modifying this template and before distributing it.

Poster: Jon Miller | Post Date: February 17, 2008 8:21 PM

I have question, I know how to calculate tt, my problem is a I have a production line that has 3 product families produced. Product 1 volume 356/day sum of cycles 80 sec, Product 2 volume is 103/day sum of cycles 200 sec and Product 3 volume is 89/day sum of cycles 152. How do I give each product family a TT.? Staffing is 3, 25800 sec. Available Time

Poster: huntr@hearthnhome.com | Post Date: June 12, 2008 12:30 PM

If you have 3 product families you need to calculate 3 takt times. Staffing is irrelevant. I wouldn't recommend running a mixed model line unless the downstream customer needs one of each product to make a set or shipment. It is better to run one product family for a few hours until the daily demand is met, then switch to another, etc. in a daily pattern. Roughly speaking you will need 2,8480 sec (80 x 356) for Product 1, 20,600 sec (103 x 200) for Product 2 and 13,528 (89 x 152) for Product 3. The sum of cycle time (I am assuming you mean manual time) is 62,608 sec so the optimum staffing is 2.4 people.

If you have one piece flow and cross trained workers it becomes much easier since you can simply divide the work content for each product by 2 or 3 evenly and run until finished based on not the "true" takt but something called "actual takt". For example if you could balance the workload and pace the flow for Product 2 at 67 seconds per person with 3 workers you could finish your daily requirement in about 3 hours and 20 min.

Good luck!

Poster: Jon Miller | Post Date: June 13, 2008 10:05 AM

Thanks

This website really help me doing report.

Poster: wan | Post Date: February 6, 2009 5:11 AM

This is a good calculator. But it assumes that the customer demand has the same work time per piece, do you have a calculator for mixed model cells?

Poster: Ron Benedict | Post Date: April 6, 2009 11:14 AM

Hi Ron

You can use the same calculator.

If you are running lots of different models AAABBBCCC then determine the capacity needed (minutes) to run that lot, set that as the available time, and calculate the model-specific takt.

If you are running ABCABCABC then it it's a question of how balanced the cycle times are between A, B and C. The calculation of net available time divided by demand does not change because the cycle time or mix changes, unless the cycle time difference is greater than 10%.

In that case you will want to use a weighting to determine the 1-piece equivalent for the longer cycle time (not lead time but processing time) items. For example if product A was 5 minutes, B was 7.5 minutes and C was 10 minutes processing time then demand weighting for item A = 1, item B = 1.5 and C = 2. If your demand was 30 pieces of A, 30 pieces of B and 30 pieces of C, you would want to calculate demand based on 30, 45 and 60 or 135 pieces equivalent.

It's much easier to do it in lots AAABBBCCC so unless the cycle times are stable and nearly identical, I would go that way. Feel free to add more detail and I'll try to explain more precisely to your situation.

Poster: Jon Miller | Post Date: April 12, 2009 6:30 PM

Its a followup question from the previous comment. Our production facility lines have seven lines and each one of them supports more than 50 products. Each one of them has its own rate of production based on the process constraint.Both work content and demand changes for all these products. Is there any way we can have a common TAKT or a rate for each one of those models. How do we determine the correct TAKT for all these 100 models.

Thanks,

Poster: JP | Post Date: October 6, 2009 12:51 PM

Hi JP

I am not sure having a common takt is desirable in your case. Unless the models that run on each line fit into product families with similar process flows and work content, pacing the work to takt time on a high mix line will be difficult. It's a bit hard to give you a good answer while discussing an abstract scenario. If you would e-mail me some more details perhaps I can give a more precise answer.

Poster: Jon Miller | Post Date: October 6, 2009 4:34 PM

I have just moved from a role where the takt was 55 seconds into a new role where the customer requirement is one part every day. Does anyone have experience running to this kind of timescale.

Poster: Steve Coath | Post Date: August 14, 2010 12:15 PM

Hi, i have question on how to calculate takt time. A demand of product A is 1000 nos. It takes 2 shifts (8hrs) and 1 hour break to get it done. For the first shift usually takes 20 -30 minutes to warm up the machine for the initial process. How should I count the cycle time for the initial process? and how should I count the total cycle time if it involve 2 shift then?

Poster: Jason | Post Date: September 13, 2011 11:07 PM
Comments are moderated to filter spam and inappropriate content. There may be a delay before your comment is published.

Leave a comment