Also known as Set up wheel interval this is the time it takes for you run all the products in a cell. This is directly impacted by the number of SKUs and changeover times. Here is how to calculate EPEI for our cell:
# of SKUs -10
Cycletime - 2 min
Avg. Daily Demand - 20
Work Hours - 8 hrs= 480 minutes
Avg. Change Over time - 60 min
By now, you know that inventory is a reflection of you leadtime. The longer the EPEI the more you would have to store. The ultimate goal is to run every part every day. Let's start with that assumption. For the above example:
Work Hours = 480 min
Production Time Needed = 10 SKU's * 20/Day*2 min/Part = 400 minutes
Time available for set up = 480 -400 = 80 min
# of set ups in that time = 80/60 = 1.33
Our goal is to run every part every day which means 10 changeovers. This tells us that we can do only 1.33 changeovers. So it cannot be done in a day. EPEI is the time it takes to cycle through all the items, which means we have to 10 changeovers. If I can do 1.33 changeover/day, it'll take us 7.5 days to do 10. That's your EPEI 7.5 days. Your batch size would be 7.5 days worth of demand.
Let's play with the number a bit more:
You reduce cycletime by 50% to 1 min
Prod time needed= 200 min
Time avai for set up = 280
# of set ups = 280/60 = 4+
EPEI = 2.5 days (See how a 50% reduction gave your 75% reduction in EPEI and hence, 75% reduction in Inventory)
You reduce setup by 50% to 30 min
# of setups = 80/30=2.66
EPEI = 3.75
Depending on your # of SKU these might change (You can do the math by yourself).
If you are struggling with a mixed model cell, my suggestion is always to start with level loading with the existing setup times and cycletime. Of course, that means a lot of inventory at the start. But it gives you some semblance of predictability. That gives you the option of planning the time for improvements. When you start to reduce your Setup time and Cycle Time, all you need is to recalculate the EPEI and reduce your batch sizes. Eventually you'll get to your Every Part Every Day. In case, you are thinking of reducing setups and cycle time before level loading, good luck with that. It is possible but easier with doing it the other way.
This is just the first level to show you the impact of level loading. Good lean practitioners understand the philosophy and create their own systems. Don't try to blindly copy/follow what I have said so far. Think for yourself and do what is best for the situation that you are facing.
# of SKUs -10
Cycletime - 2 min
Avg. Daily Demand - 20
Work Hours - 8 hrs= 480 minutes
Avg. Change Over time - 60 min
By now, you know that inventory is a reflection of you leadtime. The longer the EPEI the more you would have to store. The ultimate goal is to run every part every day. Let's start with that assumption. For the above example:
Work Hours = 480 min
Production Time Needed = 10 SKU's * 20/Day*2 min/Part = 400 minutes
Time available for set up = 480 -400 = 80 min
# of set ups in that time = 80/60 = 1.33
Our goal is to run every part every day which means 10 changeovers. This tells us that we can do only 1.33 changeovers. So it cannot be done in a day. EPEI is the time it takes to cycle through all the items, which means we have to 10 changeovers. If I can do 1.33 changeover/day, it'll take us 7.5 days to do 10. That's your EPEI 7.5 days. Your batch size would be 7.5 days worth of demand.
Let's play with the number a bit more:
You reduce cycletime by 50% to 1 min
Prod time needed= 200 min
Time avai for set up = 280
# of set ups = 280/60 = 4+
EPEI = 2.5 days (See how a 50% reduction gave your 75% reduction in EPEI and hence, 75% reduction in Inventory)
You reduce setup by 50% to 30 min
# of setups = 80/30=2.66
EPEI = 3.75
Depending on your # of SKU these might change (You can do the math by yourself).
If you are struggling with a mixed model cell, my suggestion is always to start with level loading with the existing setup times and cycletime. Of course, that means a lot of inventory at the start. But it gives you some semblance of predictability. That gives you the option of planning the time for improvements. When you start to reduce your Setup time and Cycle Time, all you need is to recalculate the EPEI and reduce your batch sizes. Eventually you'll get to your Every Part Every Day. In case, you are thinking of reducing setups and cycle time before level loading, good luck with that. It is possible but easier with doing it the other way.
This is just the first level to show you the impact of level loading. Good lean practitioners understand the philosophy and create their own systems. Don't try to blindly copy/follow what I have said so far. Think for yourself and do what is best for the situation that you are facing.