Lean Project Merged PDF

Resource Planning Session 4
ISE 454
Production Planning
and Inventory Control
Dr. Abdulrahman R. Alenezi
ISE 454 Production Planning and Inventory Control Dr. Abdulrahman Alenezi
1
Business Long-Term
Planning Strategic
(Aggregate) Resource Short-Term
Demand
Production
Management
Planning Planning Strategic
Master Rough-Cut
Production Capacity
Scheduling Planning
Tactical
Material Capacity
Purchasing Requirements Requirements
Planning Planning
Shop
Detailed
Floor
Sequencing
Operational
Control
Manufacturing Operational
2
• Resource Planning
1. Facility location: is the process of identifying the best geographic
location for a service or production facility
2. Capacity planning: is the process of establishing the output rate

that may be needed at a facility
3
1. Facility Location
General factors in selecting the Facility Location
• Market (customer) proximity
– High population areas
• Suppliers proximity
– Transportation costs
• Labor proximity
– Proximity—low salary rates, special skills availability
• Competitor proximity
– Clustering—due to a major resource in the area).
4
1. Facility Location
• Location Analysis Methods
– Analysis should follow 3 step process:
• Step 1: Identify main location factors
• Step 2: Develop location alternatives
• Step 3: Evaluate locations alternatives
1. Factor rating method
2. Load-distance model
3. Center of gravity approach
4. Economic Analysis.
5. Network Optimization Models
5

• Six steps:
1. Develop a list of relevant factors.
2. Assign a weight to each factor reflecting its relative
importance to the firm.
3. Develop a rating scale for the factors.
4. Score each location on each factor based on the scale.
5. Multiply the scores by the weights for each factor and
total the weighted scores for each location.
6. Make a recommendation based on the maximum point
score, considering other [quantitative?] factors.
6
7
• Calculate the rectilinear distance:
dAB  30 10  40 15  45 miles
• Example: Matrix Manufacturing is considering where to

locate its warehouse in order to service its four Ohio stores
located in Cleveland, Cincinnati, Columbus, Dayton. Two sites
are being considered; Mansfield and Springfield, Ohio. Use
the load-distance model to make the decision.
8
• Multiply by the number of

loads between each site
and the four cities
9
Computing the Load-Distance Score for Springfield
City Load Distance Load-Distance
Cleveland 15 20.5 307.5
Columbus 10 4.5 45
Cincinnati 12 7.5 90
Dayton 4 3.5 14
Total Load-Distance Score 456.5
10
Computing the Load-Distance Score for Mansfield
City Load Distance Load-Distance
Cleveland 15 8 120
Columbus 10 8 80
Cincinnati 12 20 240
Dayton 4 16 64
Total Load-Distance Score 504
11
• The load-distance score for Mansfield is higher than for

Springfield. The warehouse should be located in Springfield.
12

• Place the locations to be supported on a coordinate system
(like a graph).
1. Calculate the center of gravity:
xcg  Coordinate 
 li xi
ycg  Coordinate 
 li yi
Where:
 li  li
xi = x-coordinate of location i.
yi = y-coordinate of location i.
li = quantity (load) of goods moved to/from location i.
13

• Example: Matrix Manufacturing is considering where to
locate its warehouse in order to service its four Ohio stores
located in Cleveland, Cincinnati, Columbus, Dayton. Use the
Center of gravity approach to make the decision.
14

Computing the center of gravity for Matrix Manufacturing
Location (X,Y) Load (li) li xi li yi
Cleveland (11,22) 15 165 330
Columbus (10,7) 10 100 70
Cincinnati (4,1) 12 48 12
Dayton (3,6) 4 12 24
Total 41 325 436
• Computing the Center of Gravity for Matrix Manufacturing

xcg 
 li xi 325
  7.93 ycg 
 li xi 436
  10.63
 li 41  li 41
Is there another possible warehouse location closer to the C.G.
that should be considered?? Why?
15
4. Locational Break-Even Analysis Example

Three locations:
Fixed Variable Total

City Cost Cost Cost
Cincinnati $30,000 $75 $180,000
Cleveland $60,000 $45 $150,000
Chicago $110,000 $25 $160,000
Selling price = $120
Expected volume = 2,000 units
Total Cost = Fixed Cost + Variable Cost x Volume

16
4. Locational Break-Even Analysis Example

–
$180,000 –
–
$160,000 –
$150,000 –
–
$130,000 –
–
Annual cost
$110,000 –
–
–
$80,000 –
–
$60,000 –
–
–
Cincinnati Chicago
$30,000 – lowest
Cleveland lowest
lowest
– cost
cost
cost
$10,000 –
| | | | | | |
–
0 500 1,000 1,500 2,000 2,500 3,000
Volume
17

• Allocating demand to production facilities
• Locating facilities and allocating capacity
Key Costs:
• Fixed facility cost

• Transportation cost
• Production cost
• Inventory cost
• Coordination cost
Which plants to establish? How to configure the network?
18
Materials Customer
Vendor Finished Customer Store
DC DC
DC Goods DC
Customer
Component Store
Vendor Manufacturin
DC g Plant Customer Customer
Warehouse DC Store
Components
DC Customer
Vendor Store
DC Finished
Customer
Goods DC
Final DC Customer
Assembly Store
19

Demand Allocation Model
• Which market is served by which plant?
• Which supply sources are used by a plant?
xij = Quantity shipped from plant site i to customer j
n m
Min  cij xij
i 1 j 1
s.t.
n
x  D
i 1
ij j
x  K
j 1
ij i
x ij
0
20

Plant Location with Multiple Sourcing
• yi = 1 if plant is located at site i, 0 otherwise
• xij = Quantity shipped from plant site i to
customer j
n n m
Min  f y   c x ij ij
i i
i 1 i 1 j 1
s.t.
n
x  D
i 1
ij j
x  K y
j 1
ij i i
 y  k ; y {0,1}
i 1
i i
21

Set Covering Models
Define:
cj cost of locating facility at site j
aij = { 1
0
if facility located at site j can cover customer i
Otherwise
xj = { 1
0
if facility located at site j
Otherwise
The set covering problem is to:
22

Set Covering Models
The set covering problem is to:
n
Minimize c x
j 1
j j
s.t.
n
a
j 1
i, j x j  1, i  1..n
x j  0,1 , j  1..n
23

Set Covering Models
Greedy Heuristic for Set Covering Problem:
Step 1: If cj = 0, for any j = 1, 2, ..., n, set xj = 1 and remove all constraints
in which xj appears with a coefficient of +1.
Step 2: If cj > 0, for any j = 1, 2, ..., n and xj does not appear with a +1
coefficient in any of the remaining constraints, set xj = 0.
Step 3: For each of the remaining variables, determine cj/dj, where dj is
the number of constraints in which xj appears with a +1
coefficient. Select the variable k for which ck/dk is minimum, set
xk = 1 and remove all constraints in which xj appears with a +1
coefficient. Examine the resulting model.
Step 4 If there are no more constraints, set all the remaining variables
to 0 and stop. Otherwise go to step 1.
24

Set Covering Models
Example:
A rural county administration wants to locate several medical emergency
response units so that it can respond to calls within the county within eight
minutes of the call. The county is divided into seven population zones. The
distance between the centers of each pair of zones is known and is given in the
matrix below.
Imagine that the one that has to make the decision does not want to place a
emergency unit on B or D
25

Set Covering Models
Example:
1 2 3 4 5 6 7
1 0 4 12 6 15 10 8
2 8 0 15 60 7 2 3
[dij]= 3 50 13 0 8 6 5 9
4 9 11 8 0 9 10 3
5 50 8 4 10 0 2 27
6 30 5 7 9 3 0 27
7 8 5 9 7 25 27 0
26

Set Covering Models
Example:
The response units can be located in the center of population zones 1 through 7
at a cost (in hundreds of thousands of dollars) of 100, 80, 120 110, 90, 90, and
110 respectively. Assuming the average travel speed during an emergency to
be 60 miles per hour, formulate an appropriate set covering model to
determine where the units are to be located and how the population zones are
to be covered and solve the model using the greedy heuristic.
27

Set Covering Models
Solution:
Defining
1 if zone i’s center can be reached from center of zone j within 8 minutes
aij = { 0 otherwise
and noting that dij > 8, dij <= 8 would yield aij values of 0, 1,
respectively the following [aij] matrix can be set up.
28

Set Covering Models
Solution:
Minimize Subject to:
100 x1+80 x2+120 x3+110 x4+90 x5+90 x6+110 x7

x1 + x2 + x4 + x7 =1
x1 + x2 + x5 + x6 + x7 =1
x3 + x4 + x5 + x6 =1
x3 + x4 + x7 =1
x2 + x3 + x5 + x6 =1
x2 + x3 + x5 + x6 =1
x1 + x2 + x4 + x7 =1
x1 , x2 , x3 , x4 , x5 , x6 , x7 {0,1}
29

Set Covering Models
Solution:
Step 1: Since each cj > 0, j = 1, 2, ..., 7, go to step 2.
Step 2: Since xj appears in each constraint with a +1

coefficient, go to step 3.
30

Set Covering Models
Solution:
Step 3: c1 100 c5 90
= = 33.3 = = 22.5
d1 3 d5 4
c2 80 c6 90
= = 16 = = 22.5
d2 5 d6 4
c3 120 c7 110
= = 30 = = 27.5
d3 5 d7 4
c4 110
= = 27.5
d4 4
31

Set Covering Models
Solution:
Since the minimum ck/dk occurs for k = 2, set x2 = 1 and remove the
first two and the last three constraints. The resulting model
is shown below.

100x1+120x3+110x4+90x5+90x6+110x7
x3 + x4 + x5 + x6 =1
x3 + x4 + x7 =1
x1 , x3 , x4 , x5 , x6 , x7 {0,1}
32

Set Covering Models
Solution:
Step 4: Since we have two constraints go to step 1.

Step 1: Since c1 > 0, j = 1, 3, 4, ..., 7, go to step 2
Step 2: Since c1 > 0 and x1 does not appear in any of the
constraints with a +1 coefficient, set x1 = 0.
33

Set Covering Models
Solution:
c3 120 c7 110
Step 3: = = 60 = = 110
d3 2 d7 1
c4 110
= = 55
d4 2
c5 90
= = 90
d5 1
c6 90
= = 90
d6
1
34

Set Covering Models
Solution:
Since the minimum ck/dk occurs for k = 4, set x4 = 1 and

remove both constraints in the above model since x4 has
a +1 coefficient in each. The resulting model is shown
below.
120x3+90x5+90x6+110x7
x3 , x 5 , x 6 , x 7 =0
35

Set Covering Models
Solution:
Step 4: Since there are no constraints in the above

model, set x3 = x5 = x6 = x7 = 0 and stop.
The solution is x2 = x4 = 1; x1 = x3 = x5 = x6 = x7 =
0. Cost of locating emergency response units to
meet the eight minute response service level is
80 + 110 = 190.
36
2. Capacity planning
• Capacity is the maximum output rate of a production or service facility
• Capacity planning is the process of establishing the output rate that may
be needed at a facility:
– Capacity is usually purchased in “chunks”
– Strategic issues: how much and when to spend capital for additional
facility & equipment
– Tactical issues: workforce & inventory levels, & day-to-day use of
equipment
37
• Measuring Capacity Examples
– There is no one best way to measure capacity
– Output measures like barrels per day are easier to understand
– With multiple products, inputs measures work better
Input Measures of Output Measures
Type of Business
Capacity of Capacity
Car manufacturer Labor hours Cars per shift
Hospital Available beds Patients per month
Pizza parlor Labor hours Pizzas per day
Floor space in
Retail store Revenue per foot
square feet
38
• Capacity Information Needed
– Design capacity:
• Maximum output rate under ideal conditions
• Example: A bakery can make 30 custom cakes per day when
pushed at holiday time
– Effective capacity:
• Maximum output rate under normal (realistic) conditions
• Example: On the average this bakery can make 20 custom cakes
per day
39
• Calculating Capacity Utilization
– Measures how much of the available capacity is actually
being used:
Utilizatio n 
actual output rate
100%
capacity
– Measures effectiveness
– Use either effective or design capacity in denominator
40
• Example of Computing Capacity Utilization: In the bakery example the
design capacity is 30 custom cakes per day. Currently the bakery is
producing 28 cakes per day. What is the bakery’s capacity utilization
relative to both design and effective capacity?
actual output 28
Utilizatio n effective  (100%)  (100%)  140%
effective capacity 20
actual output 28
Utilizatio n design  (100%)  (100%)  93%
design capacity 30
41
– The current utilization is only slightly below its design capacity and
considerably above its effective capacity
– The bakery can only operate at this level for a short period of time
42
• Two General Approaches to Expanding Long-Range
Capacity
– All at Once – build the ultimate facility now and grow into
it
– Incrementally – build incrementally as capacity demand
grows
43
– Minimum Cost, Capacity-Feasible Line
• Consider a four-station line with a throughput target of two and
one-half jobs per hour or 60 jobs per day (running three shifts per
day). Assume that only one type of machine is available at each
station (although we allowed to choose the number machines to
install at each station). Let
• te = mean effective process time for machine
• m = number of identical machine at station
• k = cost per machine
• A = fixed cost of machine option
• TH = required throughput.
• ra = arrival rate equal to TH when utilization < 1
44
Fixed Cost Unit cost te
Station
($000) ($000) (hours)
1 225 100 1.50
2 150 155 0.78
3 200 90 1.10
4 250 130 1.60
• Total cost of installing the option is A+km

ra te
• WorkStation Utilization u and u 1
m
45
• For the first station
ra te
 1  m  ra te  2.5 1.50
m
m  3.75  m  4
• Total cost = A+km = 225+4*100 = $625
ra te 2.5 1.5
• Utilization = u   0.94
m 4
46
Number of Cost
Station Utilization
Machines ($000)
1 4 0.94 625
2 2 0.98 460
3 3 0.92 470
4 5 0.80 900
Total 2,455
47
Demand Management Session 5
ISE 454
Production Planning
1
Business Long-Term
Planning Strategic
Demand
Production
Management
Master Rough-Cut
Production Capacity
Scheduling Planning
Tactical
Material Capacity
Planning Planning
Shop
Detailed
Floor
Sequencing
Operational
Control
2
Demand Management
• Forecast
• What is forecasting?
Main Function is to Predict the Future
• Why are we interested?
Affects the decisions we make today
• Examples: who uses forecasting in their jobs?
Forecast demand for products and services
Forecast availability of manpower
Forecast inventory and materiel needs
3
• Characteristics of Forecasts
1. They are usually wrong!
2. A good forecast is more than a single number
• mean and standard deviation
• range (high and low)
3. Aggregate forecasts are usually more accurate
4. Accuracy decreases as we go further into the future.
5. Forecasts should not eliminate known information.
4
• Problem Characteristics
– The main characteristics of a forecasting problem are:
• Time horizon
• Level of detail
• Accuracy needed
5
• Forecast Horizons in Operation Planning
6
– Example 1:
• In production system, long-term decisions such as
opening new plats or adding capacity depend on a
forecast of demand.
• Time horizon: Long-term

• Level of detail: Not interested in individual
products, but in overall volume
• Accuracy needed: Long-term decisions do not

require exact forecast
7
– Example 2:
• In production system, mid-range decisions such as
assigning plat capacity to groups of products depend
on a forecast of demand.
• Time horizon: Intermediate

• Level of detail: Not interested in individual
products, but for group items
• Accuracy needed: Intermediate decisions require

greater accuracy
8
– Example 3:
• In production system, Short-term decisions such as
how mach product we should produce.
• Time horizon: Short-term

• Level of detail: Interested in individual products
• Accuracy needed: Short-term decisions require
very accurate forecast
9
• Forecasting
– Qualitative approaches
• No past sales data is available
• Appropriate for forecasting technology change, new product
sales, consumer tastes change, etc.
– Quantitative approaches
• Sales history is available
• Product design is stable
10
• Forecasting Methods
– Qualitative Approaches
• Market Survey
• Expert Opinion
• Delphi Technique
– Individual opinions are collected and reconsidered. Repeat until
and overall group agreement is (hopefully) reached.
– Quantitative Approaches
• Causal Models
• Time Series Methods
11
Causal Models
Let Y be the quantity to be forecasted and (X1, X2, . . . , Xn)
be n variables that have predictive power for Y.
A causal model is Y = f (X1, X2, . . . , Xn).
A typical relationship is a linear one. That is,

Y = a0 + a1X1 + . . . + an Xn.
12
Causal Methods: Simple Linear Regression

Y Regression
equation:
Estimate of Y from Yi = a + bXi
Dependent variable
regression equation
Actual
Value of Y
Value of X used
to estimate Y
X
Independent variable
13

Let:
Yˆ = the estimate value
Y = the actual value
Y = the average of all actual values
X = the independent variable
a = intercept of the straight line with Y axis
b = slope of the line
n = the total number of the data points
r = the coefficient determination close to 1 indicates an excellent fit
14

Then:
n X iYi   X i  Yi
b
n X i2   X i 
2
1 b
a  i
Y   Xi
n n
r2 

 Yˆi  Y 
2
 Yi  Y 
2
15

Example:
Using the data below estimate point sales (in 000 of units)
when advertising budget is $k2.3
Sales Advertising
Month (000 units) (000 $k)
1 264 2.5
2 116 1.3
3 165 1.4
4 101 1.0
5 209 2.0
Si = a + bAi
16

Solution:
i S A A^2 S*d
1 264 2.5 6.25 660
2 116 1.3 1.69 151
3 165 1.4 1.96 231
4 101 1 1 101
5 209 2 4 418
Total 855 8.2 14.9 1561
17

Solution:
n Ai Si   Ai  Si 5 1561  8.2  855

b   109.37
n Ai   Ai  5 14.0  8.2
2 2 2
a
1
 i
Y 
b
 i
X 
1
855  
109.37
8.2  8.36
n n 5 5
Sˆi  8.36  109.37 Ai
18

Solution:
If A = 2,300 the expected sales is:

S(2.3) = -8.36 + 109.37 (2.3) = $243.18
19

Solution: S
 Si  855  171
n 5
S  S  Sˆ  S 
2 2
i A S Ŝ
1 2.5 264 265.07 8649 8848
2 1.3 116 133.82 3025 1382
3 1.4 165 144.76 36 688
4 1 101 101.01 4900 4898
5 2 209 210.38 1444 1550
Total 8.2 855 855.03 18054 17368
r 
2  
ˆ
S i  S 
2

17368
 0.962 r  0.981
 Si  S 
2
18054
20

Example:
Using the data below estimate point sales (in 000 of units) in month 6
Sales
Month (000 units)
1 150
2 163
3 182
4 191
5 209
21
• Forecasting Methods
– Qualitative Approaches
• Market Survey
• Expert Opinion
• Delphi Technique
– Individual opinions are collected and reconsidered. Repeat until
and overall group agreement is (hopefully) reached.
– Quantitative Approaches
• Causal Models
22

A time series is just collection of past values of the variable
being predicted.
Goal is to isolate patterns in past data.

a. Trend
b. Seasonality
c. Cycles
d. Randomness
23

Quantity
Time
(a) Trend: Data increase or decrease, not
necessarily in a linear fashion
24
Year 1
Quantity
Year 2
| | | | | | | | | | | |
J F M A M J J A S O N D
Months
(b) Seasonal: Data show peaks and valleys
with a one-year cycle
25

Quantity
| | | | | |
1 2 3 4 5 6
Years
(c) Cyclical: Data show gradual increases and
decreases over extended periods of more than
one-year in length
26

Quantity
Time
(d) Randomness: Data cluster about a horizontal line

with no information
27

Let D1, D2, . . . Dn, . . . be the past values of the series to be
predicted (demand). If we are making a forecast in period t,
assume we have observed Dt , Dt-1 etc.
Let Ft, t + t  forecast made in period t for the demand in period t

+ t where t = 1, 2, 3, …
Then Ft -1, t is the forecast made in t-1 for t and Ft, t+1 is the
forecast made in t for t+1. (one step ahead) Use shorthand
notation Ft = Ft - 1, t .
28

The forecast error in period t, et, is the difference between the forecast for
demand in period t and the actual value of demand in t.
For one step ahead forecast: et  d t  Ft
1 T
Mean Absolute Deviation: MAD   et
T t 1
1  T et 
Mean Absolute Percentage Error: MAPE    100 

T  t 1 dt 
1 T
Mean Square Error: MSE   et2
T t 1
1 T 2
Root Mean Square Error: RMSE   et
T t 1
29

• Moving Average (MA)
• Exponential Smoothing (ES)
30
• Moving Average
Simple Moving Average is the average of the n most
recent observations. For a one-step-ahead
forecast:
Ft  d t 1  d t  2    d t  n 
1
n
31
• Moving Average
• Example: Calculate a 3-week simple moving average
forecast for part demand in weeks 4 and 5
Week Demand
1 400
2 380
3 411
4 415
5 ?
F4=(400+380+411)/3 = 397
F5=(380+411+415)/3 = 402
32
• Moving Average
Weighted Moving Average is calculate the forecast as
a weighted average of most recent observations
– Assign a larger weight to a more recent observation
– Weights are positive fractions summing up to one
Ft  w1d t 1  w2 d t  2    wn d t  n 
1
n
w 1
33
• Moving Average
• Example: Calculate a 3-week weighted moving average
forecast for part demand in weeks 4 and 5. The weights
are: w1=.70, w2=.20, w3=.10.
Week Demand
1 400
2 380
3 411
4 415
5 ?
F4 = 0.70(411) + 0.20(380) + 0.10(400) = 403.70

F5 = 0.70(415) + 0.20(411) + 0.10(380) = 410.70
34
• Moving Average
• Advantages of Moving Average Method
– Easily understood
– Easily computed
– Provides stable forecasts
• Disadvantages of Moving Average Method

– Requires saving all past N data points
– Lags behind a trend
– Ignores complex relationships in data
35
• Moving Average
36
• Moving Average Week Demand Forecast

• Example: Calculate a 3-week 1 55 --
weighted moving average 2 52 --
forecast for part demand in 3 52 --
weeks 4 to 10. The weights 4 44 ?
are: w1=.60, w2=.25, w3=.15.
5 47 ?
• Also calculate Mean Absolute 6 57 ?
Deviation, Mean Absolute 7 45 ?
Percentage Error, Mean Square
Error, and Root Mean Square Error. 8 48 ?
9 55 ?
10 50 ?
37
• Moving Average
Demand et
Week Forecast Ft Absolute Error  100 Square Error
dt et dt
1 55 -- -- -- --
2 52 -- -- -- --
3 52 -- -- -- --
4 44 52.45 8.45 19.20455 71.4025
5 47 47.2 0.2 0.425532 0.04
6 57 47 10 17.54386 100
7 45 52.55 7.55 16.77778 57.0025
8 48 48.3 0.3 0.625 0.09
9 55 48.6 6.4 11.63636 40.96
10 50 51.75 1.75 3.5 3.0625
Average 4.95 9.96 38.94
38
• Moving Average
Mean Absolute Deviation: MAD = 4.95 units of product
Mean Absolute Percentage Error: MAPE = 9.96% of demand
Mean Square Error: MSE = 38.94 units square of product
Root Mean Square Error: RMSE = 6.24 units of product
39

• Moving Average (MA)
40
• Simple Exponential Smoothing (ES)

• Specify the smoothing constant 
– It must be a positive fraction, 0<<1
• Specify the starting forecast value
– Usually F1 = D1
• Calculate the forecast using the equation
Ft+1 = Ft + (Dt – Ft)
41

• Example: Calculate a single exponential smoothing forecast
for period 5 assuming  = 0.1 and F1 = A1
Week Demand
1 400
2 380
3 411
4 415
5 ?
42

• Solution:
• Exponential Smoothing,  = 0.10
• F t +1 = Ft + (At – Ft )
• F1 = 400.00
• F2 = 400+0.10(400-400) = 400.00
• F3 = 400+0.10(380-400) = 398.00
• F4 = 398+0.10(415-398) = 399.30
• F5 = 399.30+.10(415-399.3) = 400.87
43
• Time Series Methods: Trend Influences

Quantity
Time
(a) Trend: Data increase or decrease, not
necessarily in a linear fashion
44
• Double Exponential Smoothing (ES)
Demand
Exponential
dt Smoothing
Demand
Bt
St
Time
45

• Specify the smoothing constant , and the trend
factor 
– Both are positive fraction, 0<<1 and 0<  <1
• Specify the starting forecast value
– Usually S1 = D1 and B1 = D2 - D1.
• Calculate the forecast using the equations
St = Ft + (Dt – Ft)
Bt = (St- St-1)+ (1- )Bt-1
Ft+k = St + kBt
where k is the number of step ahead
46
• Double Exponential Week Demand Forecast

Smoothing (ES) 1 116 --
• Example: Calculate a Double 2 133 ?
Exponential Smoothing 3 139 ?
forecast for part demand in 4 157 ?
weeks 2 to 8. The smoothing 5 154 ?
constant  = 0.4,  = 0.3. 6 159 ?
• Let S1 = D1 and B1 = 10 7 162 ?
8 172 ?
47

Demand
Week St Bt Forecast Ft
dt
1 116 116 10.00 --

2 133 128.80 10.84 126
3 139 139.38 10.76 140
4 157 152.89 11.59 150
5 154 160.28 10.33 164
6 159 165.97 8.94 171
7 162 169.74 7.39 175
8 172 -- -- 177
48
• Time Series Methods: Seasonal Influences
Year 1
Quantity
Year 2
| | | | | | | | | | | |
J F M A M J J A S O N D
Months
(b) Seasonal: Data show peaks and valleys
with a one-year cycle
49

Example: Consider the following data. Determine the
quarterly seasonally adjusted forecast for year 5 if
expected demand is 2,600 units
Quarter Year 1 Year 2 Year 3 Year 4
1 45 70 100 100
2 335 370 585 725
3 520 590 830 1160
4 100 170 285 215
Total 1000 1200 1800 2200
Average 250 300 450 550
50

1 45 70 100 100
2 335 370 585 725
3 520 590 830 1160
4 100 170 285 215
Total 1000 1200 1800 2200
Average 250 300 450 550
Actual Demand
Seasonal Index =
Average Demand
51

1 45/250 = 0.18 70 100 100
2 335 370 585 725
3 520 590 830 1160
4 100 170 285 215
Total 1000 1200 1800 2200
Average 250 300 450 550
45
Seasonal Index = = 0.18
250
52

1 45/250 = 0.18 70/300 = 0.23 100/450 = 0.22 100/550 = 0.18
2 335/250 = 1.34 370/300 = 1.23 585/450 = 1.30 725/550 = 1.32
3 520/250 = 2.08 590/300 = 1.97 830/450 = 1.84 1160/550 = 2.11
4 100/250 = 0.40 170/300 = 0.57 285/450 = 0.63 215/550 = 0.39
Quarter Average Seasonal Index

1 (0.18 + 0.23 + 0.22 + 0.18)/4 = 0.20
2
3
4
53

1 45/250 = 0.18 70/300 = 0.23 100/450 = 0.22 100/550 = 0.18
2 335/250 = 1.34 370/300 = 1.23 585/450 = 1.30 725/550 = 1.32
3 520/250 = 2.08 590/300 = 1.97 830/450 = 1.84 1160/550 = 2.11
4 100/250 = 0.40 170/300 = 0.57 285/450 = 0.63 215/550 = 0.39
Quarter Average Seasonal Index

1 (0.18 + 0.23 + 0.22 + 0.18)/4 = 0.20
2 (1.34 + 1.23 + 1.30 + 1.32)/4 = 1.30
3 (2.08 + 1.97 + 1.84 + 2.11)/4 = 2.00
4 (0.40 + 0.57 + 0.63 + 0.39)/4 = 0.50
54
Projected Annual Demand = 2600
Average Quarterly
Demand Demand = 2600/4 = 650
Management Session 6

1 45/250 = 0.18 70/300 = 0.23 100/450 = 0.22 100/550 = 0.18
2 335/250 = 1.34 370/300 = 1.23 585/450 = 1.30 725/550 = 1.32
3 520/250 = 2.08 590/300 = 1.97 830/450 = 1.84 1160/550 = 2.11
4 100/250 = 0.40 170/300 = 0.57 285/450 = 0.63 215/550 = 0.39
Quarter Average Seasonal Index Forecast

1 (0.18 + 0.23 + 0.22 + 0.18)/4 = 0.20 650(0.20) = 130
2 (1.34 + 1.23 + 1.30 + 1.32)/4 = 1.30
3 (2.08 + 1.97 + 1.84 + 2.11)/4 = 2.00
4 (0.40 + 0.57 + 0.63 + 0.39)/4 = 0.50
55

1 45/250 = 0.18 70/300 = 0.23 100/450 = 0.22 100/550 = 0.18
2 335/250 = 1.34 370/300 = 1.23 585/450 = 1.30 725/550 = 1.32
3 520/250 = 2.08 590/300 = 1.97 830/450 = 1.84 1160/550 = 2.11
4 100/250 = 0.40 170/300 = 0.57 285/450 = 0.63 215/550 = 0.39
Quarter Average Seasonal Index Forecast

1 (0.18 + 0.23 + 0.22 + 0.18)/4 = 0.20 650(0.20) = 130
2 (1.34 + 1.23 + 1.30 + 1.32)/4 = 1.30 650(1.30) = 845
3 (2.08 + 1.97 + 1.84 + 2.11)/4 = 2.00 650(2.00) = 1300
4 (0.40 + 0.57 + 0.63 + 0.39)/4 = 0.50 650(0.50) = 325
56
• Time Series Methods: Seasonal and Trend

Influences
Demand
| | | | | | | | | | | | | | | |
0 2 4 5 8 10 12 14 16
Period
57

Influences
• Winters Forecasting Procedure
• Let L be the season length
• With m complete seasons
• Therefore T = mL data points d1, d2, …, dT
• St the estimate for the constant term at period t
• Bt the estimate for the trend term at period t
• Ct the estimate for the seasonal component at period t
58

Influences
1. Calculate the average of each of the last two seasons of data, d m
and d m 1
( m 1) L ( m) L
 dt  dt
t ( m 1) L 1
d m1 
t  ( m  2 ) L 1
and dm 
L L
Now calculate BT d m  d m 1
BT 
L
59

Influences
2. Calculate the overall average
1 T
D   dt
T t 1
And use it and BT to calculate ST , the estimate of the constant as
T 1
ST  D  BT
2
60

Influences
3. Calculate the seasonal factors using the equation
1 m 1 d t il
Ct   t  T  L  1, T  L  2,..., T
m i 1 ST  (T  [t  iL ]) BT
• Now get Normalized factors  L 

C  Ct 
'

  Ct
t

61

Influences
4. To forecast k periods ahead, use
FT  k  ( ST  kBT )CT  k  gL
where g is the smallest integer greater or equal to k / L
62

Influences
5. When a new data point is obtained, set T = T+1 and update the
forecast parameters:
 dT 
ST      (1   )ST 1  BT 1 
 CT  L 
BT   ST  ST 1   (1   ) BT 1
 dT 
CT      (1   )CT  L
 ST 
63

Influences
Example: Consider the following data. Determine the
quarterly seasonally adjusted forecast for year 5 using the
winters method if d13= 97,  0.15 ,   0.1 and   0.2.
Quarter Year 1 Year 2 Year 3
1 60 69 84
2 234 266 310
3 163 188 212
4 50 59 64
Average 126.75 145.50 167.50
Overall Average 146.58
64

Influences
1 60 69 84
2 234 266 310
3 163 188 212
4 50 59 64
Average 126.75 145.50 167.50
1. Calculate the average of each of the last two seasons of data, d m and d m 1
d 2  145.50 d 3  167.50
167.50  145.50
Now calculate BT B12   5.5
4
65

Influences
1 60 69 84
2 234 266 310
3 163 188 212
4 50 59 64
Average 126.75 145.50 167.50
1 T
2. Calculate the overall average D   d t  146.58
T t 1
And use it and BT to calculate ST , the estimate of the constant as

T 1 12  1
S12  D  BT  146.58  5.5  176.83
2 2
66

Influences
1 60/116.33=0.52 69 84
2 234 266 310
3 163 188 212
4 50 59 64
Average 126.75 145.50 167.50

d1,1 60 60
C1,1     0.52
ST  (T  [1]) BT 176.83  (12  1)  5.5 116.33
67

Influences
1 0.52 69 84
2 234/121.33=1.92 266 310
3 163 188 212
4 50 59 64
Average 126.75 145.50 167.50

d 2,1 234 234
C2,1     1.92
ST  (T  [2]) BT 176.83  (12  2)  5.5 121.33
68

Influences
1 0.52 0.50 0.52
2 1.92 1.85 1.88
3 1.28 1.26 1.26
4 0.38 0.38 0.36
69

Influences
1 0.52 0.50 0.52
2 1.92 1.85 1.88
3 1.28 1.26 1.26
4 0.38 0.38 0.36
C1,1  C1, 2  C1,3
0.52  0.50  0.52
C1    0.51
3 3
C  C2, 2  C2,3 1.92  1.85  1.88
C2  2,1   1.88
3 3  Ct  4.02
C  C3, 2  C3,3 1.28  1.26  1.26
C3  3,1   1.26
3 3
C4,1  C4, 2  C4,3 0.38  0.38  0.36
C4    0.37
3 3
70

Influences
1 0.52 0.50 0.52
2 0.92 1.85 1.88
3 1.28 1.26 1.26
4 0.38 0.38 0.36
L 4 L 4
C 
'
C1  0.51  0.51 C3'  C3  1.26  1.26
1
 Ct 4.02  Ct 4.02
L 4 L 4
C  '
C2  1.88  1.87 C4'  C4  0.37  0.37
 t
2
C 4.02  Ct 4.02
71

Influences
1 0.52 0.50 0.52
2 0.92 1.85 1.88
3 1.28 1.26 1.26
4 0.38 0.38 0.36
F121  ( S12  [1]B12 )C1'
F13  ( S12  B12 )C1'
F13  (176.83  5.5)  0.51  92.99
72

Influences
5. When a new data point is obtained, set T = T+1 and update the
forecast parameters:
 d13   97 
S13    '   (1   )S12  B12   0.15   (1  0.15)176.83  5.5  183.51
 C1   0.51 
B13   S13  S12   (1   ) B12  0.1 183.51  176.83  (1  0.1)  5.5  5.62
 d13   97 

C13    
  (1   )C1  0.2  
'
  (1  0.2)  0.51  0.51
 S13   183.51 
73

Influences
1 0.52 0.50 0.52
2 0.92 1.85 1.88
3 1.28 1.26 1.26
4 0.38 0.38 0.36
F131  ( S13  [1]B13 )C2'
F14  ( S13  B13 )C2'
F14  (183.51  5.62) 1.87  353.67
74
Aggregate Production Planning (AP) Session 7
ISE 454
Production Planning
1
Business Long-Term
Planning Strategic
Demand
Production
Management
Master Rough-Cut
Production Capacity
Scheduling Planning
Tactical
Material Capacity
Planning Planning
Shop
Detailed
Floor
Sequencing
Operational
Control
2
• Why Aggregate Planning Is Necessary
– Fully load facilities and minimize overloading and

underloading
– Make sure enough capacity available to satisfy expected
demand
– Plan for systematic change of production capacity to meet
the expected customer demand
– Get the most output for the amount of resources available
3
• Inputs:
– A forecast of aggregate demand covering the selected

planning horizon (6-18 months)
– The alternative ways available to adjust short- to medium-

term capacity, to what extent each alternative could
impact capacity and the related costs
– The current status of the system in terms of workforce

level, inventory level and production rate
4
• Outputs:
– A production plan: aggregate decisions for each period in

the planning horizon about
• Workforce level
• Inventory level
• Production rate
– Projected costs if the production plan was implemented
5
• Medium-Term Capacity Adjustments

– Workforce level
• Hire or layoff full-time workers
• Hire or layoff part-time workers
• Hire or layoff contract workers
– Utilization of the work force
• Overtime
• Idle time (undertime)
• Reduce hours worked
– Inventory level
• Finished goods inventory
• Backorders/lost sales
– Subcontract
6
• Aggregation Standard
Hours
Monthly
Production
Product Demand for
– Must define a common unit Required to
Demand
Hours
Produce
of measure to aggregate
A1 5 200 1,000
demand and for different
products and current A2 2.5 100 250
inventory. A3 0.75 1,000 750
• Time Product
1,300 2,000
Family A
• Money
– Capacity must be measured Monthly
Standard Unit Production
in the same unit as the Product
Cost
Demand for
Cost ($)
aggregate product Demand
• Time (natural) A1 12 200 2,400
• Money (use standard labor A2 8 100 800

rates and equipment costs) A3 4 1,000 4,000
Product
1,300 7,200
Family A
7
• Basic Manufacturing Equation

Beginning Inventory I(t-1) 1,000
+ Production Q(t) 3,500
= Available (Net) Inventory to Ship N(t) 4,500
- Shipments and Sales D(t) 4,000
=Ending Inventory I(t) 500
8
• Strategies for Balancing Supply and Demand

– Zero Inventory: Adjust capacity to match demand
– Level Production: Use Inventory to absorb fluctuation in

demand.
– Mixed: Some combination of the Zero Inventory and

Constant Production strategy.
– Optimization: Math programming model.
9
• Simple Spreadsheet Strategies for Balancing Supply

and Demand
• Capacity (Production) in each time period is varied to
exactly match the forecasted aggregate demand in that
time period
• Capacity is varied by changing the workforce level
• Finished-goods inventories are minimal
• Labor and materials costs tend to be high due to the
frequent changes
10

• Hire and fire workers
• Flexible Workforce: increase and decrease working (over & under-
time)
• Subcontract work to other firms
• Use part-time workers
Demand
Units
Production
Time
11
• Zero Inventory:
Standard
Hours
Product
Required to
produce
A1 2.5
A2 5
A3 0.4
Monthly Demand forecast (Units)
Product March April May June July August Total
A1 220.8 265.6 317.6 283.2 254.4 232 1574
A2 165.8 199.2 238.2 212.4 190.8 174 1180
A3 3450 4150 4962.5 4425 3975 3625 24588
12
• Zero Inventory:
Monthly Demand forecast (Units)
A1 220.8 265.6 317.6 283.2 254.4 232 1574
A2 165.6 199.2 238.2 212.4 190.8 174 1180
A3 3450 4150 4962.5 4425 3975 3625 24588
Forecast in units converted to production hours
A1 552 664 794 708 636 580 3934
A2 828 996 1191 1062 954 870 5901
A3 1380 1660 1985 1770 1590 1450 9835
Aggregate demand forecast (production hours)
March April May June July August Total
A 2760 3320 3970 3540 3180 2900 19671
13
Parameter
Days Known
Units/worker 4 Estimated 4 gears/worker-day
Demand Given (in gear)
Workers needed Decision Variable
Workers available 35 Workforce size 35 at start of horizon
Workers hired
Hiring Cost 450 $450/worker
Workers laid off
Lay off cost 600 $600/worker
Workers used
Labor cost 120 $120/worker-day
Units produced
Net inventory 0
Holding cost 5 $5/unit-month
Backorder cost 15 $15/unit-month
Total cost
Para. January February March April May June Total

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Workers available 35
Workers hired
Hiring Cost 450
Workers laid off
Lay off cost 600
Workers used
Labor cost 120
Units produced
Net inventory 0
Holding cost 5
Backorder cost 15
Total cost

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Workers available 35 35 33 42 44 43 37 234
Workers hired 0 9 2 0 0 0 11
Hiring Cost 450 0 4050 900 0 0 0 4950
Workers laid off 2 0 0 1 6 4 13
Lay off cost 600 1200 0 0 600 3600 2400 7800
Workers used 33 42 44 43 37 33 232
Labor cost 120 83160 100800 121440 108360 97680 87120 598560
Units produced 2760 3320 3970 3540 3180 2900 19670
Net inventory 0 0 0 0 0 0 0 0
Holding cost 5 0 0 0 0 0 0 0
Backorder cost 15 0 0 0 0 0 0 0
Total cost 611310


demand.

17

demand.
• Backordering: When inventory is short, provide the service or
product at a later time period.
Demand
Units
Production
Time
18

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Workers hired
Hiring Cost 450
Workers laid off
Lay off cost 600
Workers used
Labor cost 120
Units produced
Net inventory 0
Holding cost 5
Backorder cost 15
Total cost

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Hiring Cost 450 1800 0 0 0 0 0 1800
Lay off cost 600 0 0 0 0 0 0 0
Workers used 39 39 39 39 39 39 234
Labor cost 120 98280 93600 107640 98280 102960 102960 603720
Units produced 3276 3120 3588 3276 3432 3432 20124
Net inventory 0 516 316 -66 -330 -78 454
Holding cost 5 2580 1580 0 0 0 2270 6430
Backorder cost 15 0 0 990 4950 1170 0 7110
Total cost 619060

– Level Production: Constant number of workers needed
to eliminate backorders
Para. January February March April May June
Days 21 20 23 21 22 22
Sum Days
Demand 2760 3320 3970 3540 3180 2900
Sum Demand
Workers 4
21

– Level Production: Constant number of workers needed
to eliminate backorders
Para. January February March April May June
Days 21 20 23 21 22 22
Sum Days 21 41 64 85 107 129
Demand 2760 3320 3970 3540 3180 2900
Sum Demand 2760 6080 10050 13590 16770 19670
Workers 4 33 38 40 40 40 39
22

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Workers hired
Hiring Cost 450
Workers laid off
Lay off cost 600
Workers used
Labor cost 120
Units produced
Net inventory 0
Holding cost 5
Backorder cost 15
Total cost

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Hiring Cost 450 2250 0 0 0 0 0 2250
Lay off cost 600 0 0 0 0 0 0 0
Workers used 40 40 40 40 40 40 240
Labor cost 120 100800 96000 110400 100800 105600 105600 619200
Units produced 3360 3200 3680 3360 3520 3520 20640
Net inventory 0 600 480 190 10 350 970
Holding cost 5 3000 2400 950 50 1750 4850 13000
Backorder cost 15 0 0 0 0 0 0 0
Total cost 634450


demand.

25

Days 21 20 23 21 22 22 129
Units/worker 4 84 80 92 84 88 88 516
Demand 2760 3320 3970 3540 3180 2900 19670
Workers needed 33 42 44 43 37 33 232
Hiring Cost 450 1350 0 1800 0 0 0 3150
Lay off cost 600 0 0 0 0 4200 0 4200
Workers used 38 38 42 42 35 35 230
Labor cost 120 95760 91200 115920 105840 92400 92400 593520
Units produced 3192 3040 3864 3528 3080 3080 19784
Net inventory 0 432 152 46 34 -66 114
Holding cost 5 2160 760 230 170 0 570 3890
Backorder cost 15 0 0 0 0 990 0 990
Total cost 605750
Zero-Inventory Level/BO Level/NoBO Mixed
Hiring Cost 4,950 1,800 2,250 3,150

Lay off cost 7,800 0 0 4,200
Labor cost 598,560 603,720 619,200 593,520
Holding cost 0 6,430 13,000 3,890
Backorder cost 0 7110 0 990
Total cost 611,310 619,060 634,450 605,750
Workers at end 33 39 40 35
Inventory at end 0 454 970 114
• Spreadsheet Methods
– Disadvantage
• Trial-and-error not optimization
– Advantage
• Planner can adjust the solution based on experience
• Planner can understand how the solution was created – can easily
see the impact of a decision and can easily change it.
– Based on the setting, advantages may or may not be

more important than disadvantages
• Manager will not implement something they do not
trust/understand
28


demand.

29
• Reminder: Standard LP Model

Max/Min z= c1x1 + c2x2 + … + cnxn
Subject to: a11x1 + a12x2 + … + a1nxn ≤ b1 (or ≥, =)
a21x1 + a22x2 + … + a2nxn ≤ b2
:
am1x1 + am2x2 + … + amxn ≤ bm
xj = decision variables
bi = constraint levels
cj = objective function coefficients
aij = constraint coefficients
30
• Basic Aggregate Planning Model

Parameters
T = planning horizon length (periods)
t = index of periods, t=1, 2, …, T
D(t) = forecasted number of units demanded in period t
n(t) = number of units that can be made by one worker in period t
CP(t) = variable cost to produce one unit in period t
CW(t) = cost to employ one worker in period t
CH(t) = cost to hire one worker in period t
CL(t) = cost to layoff one worker in period t
CI(t) = cost to carry one unit of inventory from period t to t+1
CB(t) = cost to backorder one unit in period t
31
• Basic Aggregate Planning Model

Decision Variables
P(t) = number of units produced during period t
W(t) = number of workers employed during period t
H(t) = number of workers hired at the start of period t
L(t) = number of workers laid off at the start of period t
I(t) = number of units held in inventory at the end of period t
B(t) = number of units backoredered at the end of period t
32
• Constraints
Capacity
P(t) ≤ n(t) × W(t) t = 1, 2, …, T
Workforce Balance
W(t) = W(t-1) + H(t) – L(t) t = 1, 2, …, T
W(0) = Initial workforce size
Inventory Balance
Net Inventory = I(t) – B(t)
At least one must be zero for each t ( LP model will force this to occur)
I(t) – B(t) = I(t-1) - B(t-1) + P(t) – D(t) t = 1, 2, …, T

I(0) = Initial Inventory
B(0) = Initial Backorders
33
• Constraints
Maximum Inventory / Storage Capacity
I(t) ≤ Storage capacity in period t t = 1, 2, …, T
Maximum Final Inventory

I(T) ≤ Maximum ending inventory
Rate of Change in Workforce – no more than 10%

L(t) ≤ 0.10 × W(t) t = 1, 2, …, T
H(t) ≤ 0.10 × W(t) t = 1, 2, …, T
34
• Costs
Production CP(t) × P(t)
Employment CW(t) × W(t)
Hiring CH(t) × H(t)
Layoff CL(t) × L(t)
Inventory CI(t) × I(t)
Backorder CB(t) × B(t)
Total
 CPt  Pt  CWt Wt  CH t  H t  CLt  Lt  CIt  I t  CBt  Bt 

T
t 1
35
• LP Model
Minimize
 CPt  Pt  CWt Wt  CH t  H t  CLt  Lt  CIt  I t  CBt  Bt 

T
t 1
Subject to
Pt  nt  Wt t  1,2,T
Wt  Wt 1  H t  Lt t  1,2,T
I t  Bt  I t 1  Bt 1  Pt  Dt t  1,2,T
Pt ,Wt , H t , Lt , I t  0 t  1,2,T
36
• Example – No Backorders
• Parameters
T =4
D = 80000, 50000, 120000, 150000
N = 1000 units/worker-period
CW = $1000, $1050, $1100, $1150
CH = $100
CL = $500
CI = $0.50/unit-period
• Backorders (and shortages) not allowed
• CP can be ignored, since production costs do not change from period t = 1

to t = T (no advantage to produce in one period vs. another, and total
quantity to produce is not a decision variable).
37
• LP Model
Min Z  $100H1  H 2  H 3  H 4   $500L1  L2  L3  L4   $1000  W1
$1050  W2  $1100  W3  $1150  W4  $0.50I1  I 2  I 3  I 4 
Subject to
P1 – I1 = 80,000 (1) Inventory
where
I1 + P2 – I2 = 50,000 (2) Constraint
I2 + P3 – I3 = 120,000 (3) Ht = #hired for period t
I3 + P4 – I4 = 150,000 (4) Ft = #fired for period t
P1 – 1,000W1 = 0 (5) Capacity
It = inventory at end of
P2 – 1,000W2 = 0 (6) Constraint period t
P3 – 1,000W3 = 0 (7)
P4 – 1,000W4 = 0 (8)
W1 - H1 + F1 = 0 (9) Workforce
W2 –W1 - H2 + F2 = 0 (10) Constraint
W3 –W2 - H3 + F3 = 0 (11)
W4 –W3 - H4 + F4 = 0 (12)
38
Inventory Management Session 9
ISE 454
Production Planning
1
• Inventory is a stock of items held to meet future demand. It

includes:
– Raw materials
– Finished products
– Component parts
– Supplies
– Work in process
2
• Independent and Dependent Demand Inventory:

– Independent demand
• Items demanded by external customers (Kitchen Tables)
– Dependent demand
• Items used to produce final products (table top, legs, hardware,
paint, etc.)
• Demand determined once we know the type and number of final
products
3
• Independent and Dependent Demand Inventory:

– Independent demand
• Uncertain / forecasted
• Continuous Review / Periodic Review
– Dependent demand
• “Requirements” / planned
• Materials Requirements Planning / Just in Time
4
• Purposes of Inventory :
• Meet variations in customer demand:
– Meet unexpected demand
– Smooth seasonal demand
• Pricing related:
– Temporary price discounts
– Hedge against price increases
– Take advantage of quantity discounts
• Process & supply surprises
– Internal – To maintain independence of operations
– External – To provide a buffer for variation in raw material delivery
time
5
• Reasons To Not Hold Inventory :

– Carrying cost
• Cost of storage
– Takes up valuable factory space
• Especially for in-process inventory
– Inventory covers up “problems” …
• That are best exposed and solved
6
• Inventory Hides Inventory:
Bad
Design
Lengthy Poor
Setups Quality
Machine
Inefficient Unreliable
Breakdown
Layout Supplier
7
• To Expose Problems: Reduce Inventory Levels
Bad
Design
Lengthy Poor
Setups Quality
Machine
Inefficient Unreliable
Breakdown
Layout Supplier
8
• Remove Sources of Problems
Poor
Quality
Lengthy
Setups
Bad
Machine
Design Inefficient Unreliable
Breakdown
Layout Supplier
9
• Inventory Costs :
– Purchasing cost (c ): is the per-item cost paid to the supplier .
– Ordering cost (A): the cost of preparing and monitoring the order- is
occur each time an order is placed with supplier.
Total cost = A + cQ
where Q is ordering Quantity.
– Holding cost includes the following costs:

• Interest or opportunity
• Storage and handling
• Taxes and insurance
• Damage
10
• Types of Inventory :
– Cycle inventory
• Average cycle inventory = (Q+0)/2
– Safety stock inventory
– Pipeline inventory = DL = d*L

• d = daily demand
• L = lead time (in days)
11
• ABC Analysis
– ABC analysis divides the total inventory into 3 classes A,B
& C as follows:
• ‘A’ items: 20% of SKUs, 80% of dollars
• ‘B’ items: 30 % of SKUs, 15% of dollars
• ‘C’ items: 50 % of SKUs, 5% of dollars
– ABC analysis provides a mechanism for identifying items

which will have a significant impact on overall inventory
cost
– Percentages are approximated.
12
100 — +Class C
+Class B
90 —
Class A
80 —
Percentage of dollar value
70 —
60 —
50 —
40 —
30 —
20 —
10 —
0—
10 20 30 40 50 60 70 80 90 100
Percentage of items
13
Percentage
Annual of Total
Usage in Dollar Dollar
Item Units Unit Cost Usage Usage
1 5,000 $ 1.50 $ 7,500 2.9%
2 1,500 8.00 12,000 4.7%
3 10,000 10.50 105,000 41.2%
4 6,000 2.00 12,000 4.7%
5 7,500 0.50 3,750 1.5%
6 6,000 13.60 81,600 32.0%
7 5,000 0.75 3,750 1.5%
8 4,500 1.25 5,625 2.2%
9 7,000 2.50 17,500 6.9%
10 3,000 2.00 6,000 2.4%
Total $ 254,725 100.0%
14
Sort based on the Percentage of total dollar usage
Cumulative
Perc. of Cumulative Perc. of Perc. of
Annual Total Perc. of Annual Annual
Usage in Dollar Dollar Total Dollar Usage Usage in
Class Item Units Unit Cost Usage Usage Usage in Units Units
A 3 10,000 10.50 105,000 41.2% 41.2% 18.0% 18.0%
A 6 6,000 13.60 81,600 32.0% 73.3% 10.8% 28.8%
B 9 7,000 2.50 17,500 6.9% 80.1% 12.6% 41.4%
B 2 1,500 8.00 12,000 4.7% 84.8% 2.7% 44.1%
B 4 6,000 2.00 12,000 4.7% 89.5% 10.8% 55.0%
C 1 5,000 $ 1.50 $ 7,500 2.9% 92.5% 9.0% 64.0%
C 10 3,000 2.00 6,000 2.4% 94.8% 5.4% 69.4%

C 8 4,500 1.25 5,625 2.2% 97.1% 8.1% 77.5%
C 5 7,500 0.50 3,750 1.5% 98.5% 13.5% 91.0%
C 7 5,000 0.75 3,750 1.5% 100.0% 9.0% 100.0%
15
45.0% 120.0%
40.0%
100.0%
Cumulative % Usage
35.0%
Percent Usage
30.0% 80.0%
25.0%
60.0%
20.0%
15.0% 40.0%
10.0%
20.0%
5.0%
0.0% 0.0%
3 6 9 2 4 1 10 8 5 7
Item No.
Percentage of Total Dollar Usage Cumulative Percentage
16
• Inventory Management Approaches

– A-items
• Track carefully (e.g. continuous review)
• Sophisticated forecasting to assure correct levels
– C-items
• Track less frequently (e.g. periodic review)
• Accept risks of too much or too little (depending on the
item)
17
• Inventory Decisions
– How much to order?
– When to order?
18
• Economic Order Quantity (EOQ) Model

• Demand rate D is constant, recurring, and known
• Amount in inventory is known at all times
• Ordering (setup) cost A per order is fixed
• Lead time L is constant and known.
• Unit cost C is constant (no quantity discounts)
• Annual carrying cost is i time the average $ value of the
inventory
• No stockouts allowed.
• Material is ordered or produced in a lot or batch and the lot
is received all at once
19

Receive Inventory depletion
order (demand rate)
Q
On-hand inventory (units)
Q Average
— cycle
2
inventory
Time
20
Q D
Annual cost (dollars)
Total cost = (H) + (A)

2 Q
Q
Holding cost = (H)
2
D
Ordering cost = (A)
Q
| | | | | | | |
50 100 150 200 250 300 350 400
Lot Size (Q)
21
– How much to order?
• Solve for when the derivative of total cost with respect
to Q
• Q = 2DA/H
– When to order?
• Order when inventory falls to the “Reorder Point-level”
R so we will just sell the last item as the new order
comes in:
• R = DL
22
A product that cost $60/unit has expected weekly
demand of 18 units. Order processing costs are
$45/order and the holding cost fraction is 25 percent
of purchase price
Determine the order quantity to minimize the total

annual cost
23
– D = (18 units/week)(52 weeks)
= 936 units/year
– H = 0.25 ($60/unit) = $15/unit/year
– A = $45/order
Q = 2DA/H = 75 units/order
TC = (Q/2)H + (D/Q)A = $1,124
24
Unit Cost C $0.45 /unit
Holding cost factor i 25% /year
Ordering cost SA $15.00 /order
Demand rate D 10000 units/year
Lead time L 0.0192 year
Solutions:
Re-order point R units (rounded)
Q = sqrt(2SD/(iC)) units (rounded)
ISE 454 Production Planning and Inventory Control Dr. bdulrahman Alenezi
25
• Economic Production Quantity (EPQ)

Is an extension Economic Order Quantity (EOQ) Model
• Assumptions:
– Only one item is involved
– Annual demand is known
– Demand rate is constant
– Demand occurs continually
– Production rate is constant
– Lead time does not vary
– No quantity discounts
26
27

Is an extension Economic Order Quantity (EOQ) Model
• Variables:
– A = setup cost
– D = demand rate
– h = holding cost
– T = cycle length
– P = production rate
– c = unit cost
28

Annual Annual
Production
TC = Carrying + Ordering +
cost
cost cost
Qh  D  D
TC = 1   + A + cD
2  P Q
2DA P
Q0 
h PD
•Time between set-ups (cycle time): Q */D years
29

• A Manufacturing Company produces 100,000 unites, 10 times per year to
meet the demand of 1,000,000 unites annually. The set-up cost is $5,000
per run. Holding cost is estimated at 10% of the manufacturing cost of $1
per tile. The production capacity of the machine is 500,000 tiles per month.
The factory is open 365 days per year.
• Determine
– Optimal production lot size
– Annual holding and setup costs
– Number of setups per year
30
• Quantity Discount Models

• A quantity discount is a reduced unit price based on
purchasing a large quantity
• Example discount schedule:
31

• Four Steps to Analyze Quantity Discount Models
1. Calculate Q* for each discount price
2. If Q* is too small to qualify for that price, adjust Q*
upward
3. Calculate total cost for each Q*
4. Select the Q* with the lowest total cost
32

• Brass Department Store Example
Sells toy cars
D = 5000 cars annually
A = $49 per order
h = $0.20 per car per year
Quantity Discount Schedule
33
• Use of Safety Stock

• Safety stock (SS) is extra inventory held to help prevent
stockouts
• Frequently demand is subject to random variability
(uncertainty)
• If demand is unusually high during lead time, a stockout will
occur if there is no safety stock
34
• Use of Safety Stock
35
• Determining Safety Stock Level

Need to know:
• Probability of demand during lead time (DDLT)
• Cost of a stockout (includes all costs directly or indirectly
associated, such as cost of a lost sale and future lost sales)
36
• Determining Safety Stock Level

ABCO Safety Stock Example
• ROP = 50 units (from previous EOQ)
• Place 6 orders per year
• Stockout cost per unit = $40
• h = $5 per unit per year
• DDLT has a discrete distribution
37
• Analyzing the Alternatives

• With uncertain DDLT this becomes a “decision making under
risk” problem
• Each of the five possible values of DDLT represents a
decision alternative for ROP
• Need to determine the economic payoff for each combination
of decision alternative (ROP) and outcome (DDLT)
38
• Stockout and Additional Carrying Costs
Additional
Stockout Cost Carrying Cost
ROP = DDLT 0 0
ROP < DDLT $40 per unit
0
short per year
ROP > DDLT $5 per unit per
0
year
39

– If the ROP = 35 units
Number of Inventory
Probability Stockout Cost Total cost
Units Cost
30 0.2 0 (35-30)*5=25 25*0.2=5
40 0.2 (40-35)*40=200 0 200*0.2=40
50 0.3 (50-35)*40=600 0 600*0.3=180
60 0.2 (60-35)*40=1000 0 1000*0.2=200
70 0.1 (70-35)*40=1400 0 1400*0.1=140
– So the total cost for this alternative is $565

40

Number of Inventory
Probability Stockout Cost Total cost
Units Cost
30 0.2 0 (45-30)*5=75 75*0.2=15
40 0.2 0 (45-40)*5=25 25*0.2=5
50 0.3 (50-45)*40=200 0 200*0.3=60
60 0.2 (60-45)*40=600 0 600*0.2=120
70 0.1 (70-45)*40=1000 0 1000*0.1=100

41

Number of
Probability Stockout Cost Inventory Cost Total cost
Units
30 0.2 0 (55-30)*5=125 125*0.2=25
40 0.2 0 (55-40)*5=75 75*0.2=15
50 0.3 0 (55-50)*5=25 25*0.3=7.5
60 0.2 (60-55)*40=200 0 200*0.2=40
70 0.1 (70-55)*40=600 0 600*0.1=60
– So the total cost for this alternative is $147.50

42

Number of
Units
30 0.2 0 (65-30)*5=175 175*0.2=35
40 0.2 0 (65-40)*5=125 125*0.2=25
50 0.3 0 (65-50)*5=75 75*0.3=22.5
60 0.2 0 (65-60)*5=25 25*0.2=5
70 0.1 (70-65)*40=200 0 200*0.1=20
– So the total cost for this alternative is $107.50

43

Number of
Units
30 0.2 0 (70-30)*5=200 200*0.2=40
40 0.2 0 (70-40)*5=150 150*0.2=30
50 0.3 0 (70-50)*5=100 100*0.3=30
60 0.2 0 (70-60)*5=50 50*0.2=10
70 0.1 (70-70)*40=0 0 0

44
ROP Total Cost

35 565
45 300
55 147.50
65 107.50
70 110
– So the best alternative where the ROP = 65 units
45
• Safety Stock With Unknown Stockout Costs

• Determining stockout costs may be difficult or impossible
• Customer dissatisfaction and possible future lost sales are
difficult to estimate
• Can use service level instead
Service level = 1 – probability of a stockout
46

Example:
• DDLT follows a normal distribution
(μ = 350, σ = 10)
• They want a 95% service level (i.e. 5% probability of a
stockout)
SS = ?
47
• Safety Stock and the Normal Distribution
48

Example:
From the standard Normal Table,
Z = 1.645 = X – 350 so X= 366.45

10
and, SS = 16.45 (which could be rounded to17)
49

Example:
• Assume a carrying cost of $1 per unit per year
• We can calculate the SS and its carrying cost for various
service levels
50

Example:
51

Example: Carrying Cost Versus Service Level
52
• Deterministic inventory model:

– Static models:
• Economic Order Quantity (EOQ)
– Dynamic models:
• Lot-for-Lot
• Fixed Order Quantity
• Wagner-Whitin model
• Stochastic inventory model
• news vendor model
• base stock and (Q,r) models
53
Dynamic Lot Sizing Notation:

t a period (e.g., day, week, month); we will consider t = 1, … ,T, where T
represents the planning horizon.
Dt demand in period t (in units)
ct unit production cost (in dollars per unit), not counting setup or
inventory costs in period t
At fixed or setup cost (in dollars) to place an order in period t
ht holding cost (in dollars) to carry a unit of inventory from period t to

period t +1
Qt the unknown size of the order or lot size in period t
54
Data: t 1 2 3 4 5 6 7 8 9 10
Dt 20 50 10 50 50 10 20 40 20 30
ct 10 10 10 10 10 10 10 10 10 10
At 100 100 100 100 100 100 100 100 100 100
ht 1 1 1 1 1 1 1 1 1 1
– Lot-for-Lot Solution
t 1 2 3 4 5 6 7 8 9 10 Total
Dt 20 50 10 50 50 10 20 40 20 30 300
Qt 20 50 10 50 50 10 20 40 20 30 300
It 0 0 0 0 0 0 0 0 0 0 0
Setup cost 100 100 100 100 100 100 100 100 100 100 1000
Holding cost 0 0 0 0 0 0 0 0 0 0 0
Total cost 100 100 100 100 100 100 100 100 100 100 1000
Since production cost c is constant, it can be

ignored.
55
Data: t 1 2 3 4 5 6 7 8 9 10
Dt 20 50 10 50 50 10 20 40 20 30
ct 10 10 10 10 10 10 10 10 10 10
At 100 100 100 100 100 100 100 100 100 100
ht 1 1 1 1 1 1 1 1 1 1
– Fixed Order Quantity Solution
t 1 2 3 4 5 6 7 8 9 10 Total
Dt 20 50 10 50 50 10 20 40 20 30 300
Qt 100 0 0 100 0 0 100 0 0 0 300
It 80 30 20 70 20 10 90 50 30 0 0
Setup cost 100 0 0 100 0 0 100 0 0 0 300
Holding cost 80 30 20 70 20 10 90 50 30 0 400
Total cost 180 30 20 170 20 10 190 50 30 0 700
56
Data: t 1 2 3 4 5 6 7 8 9 10
Dt 20 50 10 50 50 10 20 40 20 30
ct 10 10 10 10 10 10 10 10 10 10
At 100 100 100 100 100 100 100 100 100 100
ht 1 1 1 1 1 1 1 1 1 1
– Fixed Order Period Solution
t 1 2 3 4 5 6 7 8 9 10 Total
Dt 20 50 10 50 50 10 20 40 20 30 300
Qt 70 0 60 0 60 0 60 0 50 0 300
It 50 0 50 0 10 0 40 0 30 0 0
Setup cost 100 0 100 0 100 0 100 0 100 0 300
Holding cost 50 0 50 0 10 0 40 0 30 0 400
Total cost 150 0 150 0 110 0 140 0 130 0 680
57
•Wagner-Whitin model
t 1 2 3 4 5 6 7 8 9 10
Dt 20 50 10 50 50 10 20 40 20 30
ct 10 10 10 10 10 10 10 10 10 10
At 100 100 100 100 100 100 100 100 100 100
ht 1 1 1 1 1 1 1 1 1 1
Step 1: Obviously, just satisfy D1 (note we are neglecting production

cost, since it is fixed). Z1*  A1  100
j1*  1
Step 2: Two choices,  A1  h1 D2 , produce in 1
Z 2*  min  *
either j2* = 1 or j2* = 2. Z1  A2 , produce in 2
100  1(50)  150
 min 
 100  100  200
 150
j2*  1
58
• Step3: Three choices, j3* = 1, 2, 3.
 A1  h1 D2  (h1  h2 ) D3 , produce in 1

Z 3*  min Z1*  A2  h2 D3 , produce in 2
 Z*2  A3 , produce in 3
100  1(50 )  (1  1)10  170

 min 100  100  (1)10  210
150  100  250
 170
j3*  1
59
• Step 4: Four choices, j4* = 1, 2, 3, 4.
 A1  h1 D2  (h1  h2 ) D3  (h1  h2  h3 ) D4 , produce in 1
Z*  A  h D  (h  h ) D ,
 1 produce in 2
Z 4  min  *
* 2 2 3 2 3 4
 Z 2  A3  h3 D4 , produce in 3
 Z*3  A4 , produce in 4
100  1(50)  (1  1)10  (1  1  1)50  320
100  100  (1)10  (1  1)50  310

 min 
150  100  (1)50  300
170  100  270
 270
j4*  4
60
• In the Example:
– Given fact: we produce in period 4 for period 4 of a
4 period problem.
– Question: will we produce in period 3 for period 5 in
a 5 period problem?
– Answer: We would never produce in period 3 for
period 5 in a 5 period problem.
61
• Step 5: Only two choices, j5* = 4, 5.
Z 3*  A4  h4 D5 , produce in 4
Z  min  *
*
 Z 4  A5 ,
5
produce in 5
170  100  1(50)  320
 min 
270  100  370
 320
j5*  4
• Step 6: Three choices, j6* = 4, 5, 6.

–And so on.
62
• Optimal Policy:
– Produce in period 8 for 8, 9, 10 (40 + 20 + 30 = 90 units)
– Produce in period 4 for 4, 5, 6, 7 (50 + 50 + 10 + 20 = 130 units)
– Produce in period 1 for 1, 2, 3 (20 + 50 + 10 = 80 units)
t 1 2 3 4 5 6 7 8 9 10 Total
Dt 20 50 10 50 50 10 20 40 20 30 300
Qt 80 0 0 130 0 0 0 90 0 0 300
It 60 10 0 80 30 20 0 50 30 0 0
Setup cost 100 0 0 100 0 0 0 100 0 0 300
Holding cost 60 10 0 80 30 20 0 50 30 0 280
Total cost 160 10 0 180 30 20 0 150 30 0 580
63
• Deterministic inventory model:

– Static models:
• Economic Order Quantity (EOQ)
– Dynamic models:
• Lot-for-Lot
• Fixed Order Quantity
• Wagner-Whitin model
• Stochastic inventory model
• news vendor model
• base stock and (Q,r) models
64
• The Newsvendor Approach

• Assumptions:
– 1. single period
– 2. random demand with known distribution
– 3. linear overage/shortage costs
– 4. minimum expected cost criterion
• Examples:
– newspapers
– Christmas trees or other seasonal items
65
• The Newsvendor Notation

X  demand (in units), a random variable.
G ( x)  P( X  x), cumulative distributi on function of demand

(assumed continuous .)
d
g ( x)  G ( x)  density function of demand.
dx
co  cost (in dollars) per unit left over after demand is realized.
cs  cost (in dollars) per unit of shortage.
Q  production /order quantity (in units); this is the decision v ariable.
66

Units over = max {Q-X, 0}
Cost Function:
Units short = max {X-Q, 0}
Y (Q)  expected overage  expected shortage cost
 co E units over   cs E units short 
 
 co maxQ  x,0g ( x)dx  cs maxx  Q,0g ( x)dx
 
0 0
Q 
 co 
0
(Q  x) g ( x)dx  cs  ( x  Q) g ( x)dx
Q
Objective: find the value of Q that minimizes this expected cost.

67

For Y(Q): taking its derivative and setting it to 0.
To do this, we need to take the derivative of integrals with limits
that are functions of Q. A tool called Leibnitz's rule can do this.
d a2 ( Q ) a2 ( Q )  da (Q) da (Q)
dQ 
a1 (Q )
f ( x, Q)dx  
a1 (Q ) Q
[ f ( x, Q)]dx  f (a2 (Q), Q) 2
dQ
 f (a1 (Q), Q) 1
dQ
Applying this for Y(Q):

dY (Q) Q 
dQ 0 
 c0 1g ( x)dx  c s  (1) g ( x)dx  c G(Q)  c [1  G(Q)]  0
Q
0 s
cs
G(Q ) 
*
co  cs
68


G(Q )  P X  Q 
* cs
*

co  c s
Critical Ratio is that
probability stock
covers demand
G(Q*) represents the probability that demand is less than
or equal to Q*.
Note:
c o  Q *  1
cs G(x)
co  c s
c s  Q * 
Q*
69

• Scenario:
– Demand for T-shirts is exponential with mean 1000 (i.e., G(x)
= P(X  x) = 1- e-x/1000). (Note - this is an odd demand
distribution; Poisson or Normal would probably be better
modeling choices.)
– Cost of shirts is $10.
– Selling price is $15.
– Unsold shirts can be sold off at $8.
• Model Parameters:
cs = 15 – 10 = $5
co = 10 – 8 = $2
70

Solution:
Q
 cs 5
G(Q )  1  e
* 1000
   0.714
co  cs 2  5
Q*  1,253
Sensitivity: If co = $10 (i.e., shirts must be discarded) then

Q
 cs 5
G(Q )  1  e
* 1000
   0.333
co  cs 10  5
Q*  405
71

• Suppose demand is normally distributed with mean  and
standard deviation . Then the critical ratio formula reduces to:
 Q *   cs  is the cumulative distribution
G (Q * )   
   co  c s function (cdf) of the standard
normal distribution.
Q *  cs
 z where  ( z )  z is the value in the standard
 co  c s normal table.
3.00
Q*    z Note: Q* increases in both (z)

 and  if z is positive (i.e.,
if ratio is greater than 0.5).
0.00
1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103 109 115 121 127 133 139 145 151 157
0 z
72

• Scenario:
– GAP orders a particular clothing item every Friday
– mean weekly demand is 100, std dev is 25
– wholesale cost is $10, retail is $25
– holding cost has been set at $0.5 per week (to reflect
obsolescence, damage, etc.)
• Problem: how should they set order amounts?
73

• Newsvendor Parameters:
– c0 = $0.5
– cs = $15
• Solution:
15
G (Q * )   0.9677
0.5  15
 Q  100 
   0.9677
 25 
Q  100 Every Friday, they should
 1.85 order-up-to 146, that is, if
25
there are x on hand, then
Q  100  1.85(25)  146 order 146-x.
74

• Inventory is a protection against demand uncertainty.
• Amount of protection depends on “overage” and

“shortage” costs, as well as distribution of demand.
• If shortage cost exceeds overage cost, optimal order

quantity generally increases in both the mean and
standard deviation of demand.
75

Lean Project Merged PDF

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lean Project Merged PDF

Uploaded by

Copyright:

Available Formats

Resource Planning Session 4

Dr. Abdulrahman R. Alenezi

2. Capacity planning: is the process of establishing the output rate

1. Factor rating method

1. Factor rating method

dAB  30 10  40 15  45 miles

• Example: Matrix Manufacturing is considering where to

• Multiply by the number of

• The load-distance score for Mansfield is higher than for

3. Center of gravity approach

3. Center of gravity approach

3. Center of gravity approach

• Computing the Center of Gravity for Matrix Manufacturing

4. Locational Break-Even Analysis Example

Fixed Variable Total

Total Cost = Fixed Cost + Variable Cost x Volume

4. Locational Break-Even Analysis Example

5. Network Optimization Models

• Fixed facility cost

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

The set covering problem is to:

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

5. Network Optimization Models

100 x1+80 x2+120 x3+110 x4+90 x5+90 x6+110 x7

5. Network Optimization Models

Step 1: Since each cj > 0, j = 1, 2, ..., 7, go to step 2.

Step 2: Since xj appears in each constraint with a +1

5. Network Optimization Models

5. Network Optimization Models

Minimize Subject to:

5. Network Optimization Models

Step 4: Since we have two constraints go to step 1.

5. Network Optimization Models

5. Network Optimization Models

Since the minimum ck/dk occurs for k = 4, set x4 = 1 and

5. Network Optimization Models

Step 4: Since there are no constraints in the above

• Total cost of installing the option is A+km

Dr. Abdulrahman R. Alenezi

• Forecast Horizons in Operation Planning

• Time horizon: Long-term

• Accuracy needed: Long-term decisions do not

• Time horizon: Intermediate

• Accuracy needed: Intermediate decisions require

• Time horizon: Short-term

A causal model is Y = f (X1, X2, . . . , Xn).

A typical relationship is a linear one. That is,

Causal Methods: Simple Linear Regression

Causal Methods: Simple Linear Regression

Causal Methods: Simple Linear Regression

Causal Methods: Simple Linear Regression

Causal Methods: Simple Linear Regression

Causal Methods: Simple Linear Regression

n Ai Si   Ai  Si 5 1561  8.2  855

Sˆi  8.36  109.37 Ai