Supply Chain and Operations Concepts
Autor: Mikki • November 12, 2018 • 930 Words (4 Pages) • 618 Views
...
When calculating EOQ, make sure the time units all match up (weeks, months, years, days, whatever)
When calculating cost of placing orders, you have your EOQ frequency multiplied whatever your shipping costs are
o You can add in the variable costs of purchasing if you wish
When calculating annual cost to hold, calculate average inventory by dividing however much your EOQ amount is by two (EOQ / 2), then multiplying by your holding costs
o Be sure it factors out to the correct time periods
Recommendation for management (Bete Noir example), add up all costs (placing order cost, per unit costs, holding costs) and compare with the revenue gained
o Don’t forget to calculate that revenue may need a small cost taken out (such as Bete Noir when $57 had $2 subtracted for per unit packing costs)
GOODWILL
Goodwill loss at $xx per customer unable to be sold to due to shortage is factored into G, not L, because you don’t have to actually pay anything, you just prevented “losing out on the cost” so you add it to your G
o Avoidance is negative, so it increases your G
o Recalulate critical fractile
o Expected cost due to lost customer goodwill is calculated by finding expected lost sales and multiplying by the cost of goodwill per lost sale
Goodwill can also be in the L if there is a cost for having the demand and not being able to fill it, such as with “walking a customer” in a hotel example
INCREMENTAL CONTRIBUTION
Incremental contribution >> G x [1-P(Demand</=X)] – L x P(Demand</=X)
LOST SALES, EXPECTED LEFTOVER INVENTORY, EXPECTED PROFIT
When calculating to maximize expected profit and dealing with production costs and such, use the critical fractile (G and L)
When calculating expected gross margin % ((revenue – cost) / revenue), use (expected sales x price) + (expected leftover inventory x salvage value) – order (quanity x purchasing cost), all over the expected revenue portion
When calculating proability that lost sales exceed a certain number, take that number beyond the order quantity and then calculate a z-score, then check it in z-table (don’t need standard loss normal table)
QUEING, RESOURCE UTILIZATION, LITTLE’S LAW
For time in queue questions, if you get x customers arriving per hour, you can turn that into x minutes per customer, which corresponds better to the “a” (inter-arrival time)
When calculating if there is other work a resource can do, such as the re-stocking of videos, consider the utilization of the employee. Even if waits times are expected to be rather long and it appears unintuitive, you can just say “utilization is 85%, which means 15% of the time the resource can do other things”
o You can calculate other uses of time from that 1-UTIL calculation
When calculating customers waiting in line, I = R x T
o Flow rate R = 1 / a
o Flow time calculated with long formula (in formula sheet)
SERVICE LEVEL
Service level = probability. Use the z-score calculation >> z = (X – mean) / stdev
When calculating demand within %s, do z-probability < target < high z-probability (Flextronics example)
...