Inventory finish Forecasting start

1
Inventory (finish)Forecasting (start)
2
21 Inventory Forecasting

• Comments
• Continuous Review Inventory
• Variability in demand, safety stocks
• Periodic Review Inventory
• Basic definition, safety stocks
• Exercise Periodic Review Safety Stocks
• Guaranteed Exam Question (12 points)
• Supply Chain Dynamics
• Forecasting
• Time series
• Causal

3
Comments

• Cite conversations in your project reports
• Team peer evaluations (required)
• Plant visits
• April 16 (Monday) Sealy ?
• April 23 (Monday) Samaritan
• One page insights can be team effort, e-mail
  to Rich DiTieri
• Exercise make ups
• E-mail to Rich DiTieri

4
Continuous Review System

• Real demand not constant
• Assumption of no stock-out is replaced by a
  service level target
• On hand stock position is continuously
  monitored / reviewed
• As with EOQ, need to figure
• How much to order
• When to order

5
A Continuous Review System
R Reorder Point Q Order Quantity L Lead
time
If demand during lead time is higher than
anticipated, stock-outs are possible
6
Safety Stock and Service Level

• Safety stock
• buffer added to on hand inventory to cover demand
  uncertainties during lead time
• Service level
• probability that the inventory available during
  lead time will meet demand

7
Determining the Safety Stock
area under tail
m mean demand
R Reorder point
s Safety stock
The larger the value of S, the smaller the area
under the curve
8
Safety Stocks Value (continuous review)

• For 95 service level, safety stock is 1.64 x
  standard deviation of demand over the lead time
• For 98 service level it is 2 ?
• For 99.9 service level it is 3 ?

9
Continuous Review with Safety Stock Example
Average demand D-ave 600/month ordering cost
S 200/order unit cost C 0.5 holding cost factor
i 0.15 per year holding cost factor i 0.0125 per
month demand standard deviation 200 per
month Lead time L 2 months Order quantity (Q) EOQ
value
10
Continuous Review With Safety Stock
11
Continuous Review With Safety Stock
12
Continuous Review Summary

• How much to order?
• Q sqrt(2DS/iC), where D is now the average
  demand
• When to order?
• R DL safety stock
• Safety stock depends on variability in demand and
  target service level

13
Periodic Review System

• Constraint that orders are made at fixed
  intervals
• Can be internally determined
• by company periodic inventory measurement cycles
• Can be externally determined
• by suppliers delivery schedules

14
Periodic Inventory
15
A Periodic Review (P) System
16
Safety Stock Value (Periodic Review)

• Time interval covered by demand estimate
• P L
• For 95 service level, safety stock is 1.64 x
  standard deviation of demand over P L
• For 99 service level it is 2 ?
• For 99.9 service level it is 3 ?

17
When and How Much To Order (Periodic Review)

• When Every period P
• Need to cover demand over
• P L
• Assume we have Y on hand
• How Much to Order
• D x (P L) Y safety stock

18
How does Periodic Differ From Continuous?

• Less work
• because you only do inventory periodically
• More safety stock
• because the you are guessing demand over a the
  lead time and the inventory period

19
Periodic Review with Safety Stock Exercise
(guaranteed exam question)
Average demand D-ave 600/month demand standard
deviation 200 per month Lead time L 2
months Period 10 months Service level 95 Stock
on hand 1800
(a) How much should I order? (b) What is the long
term average inventory level? (Hint it is ½ the
average order size the safety stock)
20
Dynamics Of Inventory in Supply Chains

• Supply chain dynamics can wreak havoc on supply
  chain performance measures.
• Actions of downstream supply chain members can
  affect the operations of upstream members.
• The bullwhip effect increasing variability as
  you proceed upstream in the chain.
• Resolved by sharing information along the supply
  chain

21
Supply Chain Dynamics for Facial Tissue
Bullwhip Effect
Quantity ordered
Time
22
Inventory Turnover

• Average aggregate inventory value (AGV) is the
  total value of all items held in inventory for a
  firm.

• Inventory turnover is annual sales at cost
  divided by the average aggregate inventory value
  maintained for the year.

23
Forecasting
Forecasting is hard, especially when it is about
the future
ascribed to Yogi Berra
24
Examples of how Forecasts are used in Operations

• Capacity
• Facilities planning
• Workforce planning
• Production scheduling
• Inventory
• What to stock
• Process Selection and Design
• Production scale
• Technology of process equipment
• Quality
• Process yields

25
Time Frame

• Short-range (days to months)
• Procurement of materials, scheduling of jobs,
  scheduling workforce
• Medium-range (months to year)
• Hiring, adding equipment, budgeting
• Long-range (years)
• Life cycle of current products, planning new
  products, large new facilities, strategic
  planning of company goals

26
Approaches to Forecasting

• Qualitative
• (judgement-based)
• Executive Opinion
• Sales force Composite
• Delphi method (experts)
• Market survey analysis

• Quantitative
• (math model-based)
• Time series
• smoothing /averaging
• extrapolating patterns
• Causal

Advantage forces us to understand our
assumptions Disadvantage get swayed by math
based on uncertain data

• Most common method
• Can be misleading, even with lots of data and
  experience

27
Approaches to Forecasting

• Qualitative
• (judgement-based)
• Executive Opinion
• Sales force Composite
• Delphi method (experts)
• Market survey analysis

• Quantitative
• (data-based)
• Time series
• smoothing /averaging
• extrapolating patterns
• Causal

28
Time Series vs Causal

• Time series
• Based only on past behavior of the thing we are
  trying to forecast
• Find and extrapolate the pattern

• Causal
• Use other information that drives the behavior of
  the thing we are trying to forecast

29
Patterns of Demand
30
Patterns of Demand
Figure 9.1
31
Patterns of Demand
Figure 9.1
32
Patterns of Demand
Figure 9.1
33
Patterns of Demand
Figure 9.1
34
Patterns of Demand
Figure 9.1
(c) Seasonal Data consistently show peaks and
valleys.
35
Components of Time Series Data
36
Forecast Based on Patterns
37
Resultant Forecast vs Actual
Demand
Time
forecast error can in principle be reduced to the
random variation
38
Causal Forecasting

• Based on deep understanding of what is driving
  the changes in what we are forecasting
• Can be very powerful in giving accurate forecasts
• If the cause-related quantities are known or can
  be accurately forecast

39
Forecasting Attendance Based on Wins in the last
10 Games