Q: My current ‘machine efficiency’ is almost 75% (available 75% of the time). I have 15 SKUs (identical by content) in all to make and sell with a total Daily demand of about 20000 units with a standard deviation/demand variability of 0.2. The available capacity is 30000 units/day.
What will be my practical best-case capacity utilization.?
A: Someone came up with this ‘simple formula’
Capacity Utilization
=
[ 20000 x (1- 0.2)] / [30000 x 0.75 x (1 + (0.1 x ln (15))]
This works out to 55%. ln stands for natural log. So just a 10th grade math here!
Thats your real capacity utilization. Assuming simple scheduling rules. Single machine. No complexity of set ups and set up matrices
Q2: What happens when the SKU count grows to 20? (You introduced 5 new SKUs)
A2: When SKU count grows to 20, the demand variability also grows but demand doesn’t necessarily grow proportionately. Assuming daily demand to be 27000 units and variability in demand to be 0.3
[ 27000 x (1- 0.3)] / [30000 x 0.75 x (1 + (0.1 x ln (25))]
The capacity utilization will be 65%. Which means so much more competition for the resource!
Q3: What about 30 SKUs, with exactly 30000 units of daily demand and all else remaining the same as in 2?
A3: You cannot utilize more than 70% of the rated capacity. 100% is just a theoretical case. (because effective demand in numerator is less). You can work this out.
What’s the point?
1) Do not add mindless line extension SKUs and hope that you have sufficient capacity available to meet the ‘total’ demand (just because numbers seem to add up). Even if the products have identical resource needs.
2) The Delta change in capacity utilization is a highly non-linear function of delta change in SKU count.
Q4: How much capacity you can free up by retiring 10 SKUs?
A4: Now you can work it out.
Who invented this ‘formula’?
Someone by the name John Burbidge – SKU proliferation and its impact on production planning. Sometime in 1960’s. No machine learning then!
Q5: What’s in it for SAP Consultants?
A5: Resource finiteness check, Resource utilization, Capacity Profiles, Set Up times, Set up matrices, Scheduling Heuristics, Optimizers – Whatever you can do there with this knowledge.
