Demand Forecasting &
Capacity Planning
1
Demand Forecasting
2
Role of Forecasting
in a Supply Chain
• The basis for all strategic and planning decisions in a
supply chain
• Used for both push and pull processes
• Examples:
– Production: scheduling, inventory, aggregate planning
– Marketing: sales force allocation, new product introduction
– Finance: plant/equipment investment, budgetary planning
– Personnel: workforce planning, hiring, layoffs
• All of these decisions are interrelated
Characteristics of Forecasts
• Forecasts are always wrong. Should include
expected value and measure of error.
• Long-term forecasts are less accurate than
short-term forecasts (forecast horizon is
important)
• Aggregate forecasts are more accurate than
disaggregate forecasts
Types of Forecasts
Forecasting Techniques
No single method is superior
Qualitative Models: attempt Time-Series Methods: include Causal Methods: include a
to include subjective factors historical data over a time interval variety of factors
Delphi Moving Regression
Methods Average Analysis
Jury of Executive Exponential Multiple
Opinion Smoothing Regression
Sales Force Holt’s Model
Composite
Winter’s Model
Consumer
Market Survey Trend Projections
5
Qualitative Methods
Delphi Method
interactive group process consisting of obtaining information from
a group of respondents through questionnaires and surveys
Jury of Executive Opinion
obtains opinions of a small group of high-level managers in
combination with statistical models
Sales Force Composite
allows each sales person to estimate the sales for his/her region
and then compiles the data at a district or national level
Consumer Market Survey
solicits input from customers or potential customers regarding
their future purchasing plans
6
Decomposition of a Time-Series
Time series can be decomposed into:
Trend (T): gradual up or down movement over
time
Seasonality (S): pattern of fluctuations above
or below trend line that occurs every year
Cycles(C): patterns in data that occur every
several years
Random variations (R): “blips”in the data
caused by chance and unusual situations
7
Components of Decomposition
650 Actual Data
550
Trend
450
Demand
350
250
150 Cyclic
50
-50
Random
-150
0 1 2 3 4 5
Time (Years)
8
Decomposition of Time-Series
Multiplicative model assumes demand is the product of
the four components
demand = T * S * C * R
Additive model assumes demand is the summation of
the four components
demand = T + S + C + R
9
Moving Averages
Moving average methods consist of computing an
average of the most recent n data values for the time
series and using this average for the forecast of the
next period.
Simple moving average =
demand in previous ' n' periods
n
10
Three-Month Moving Average
Month Actual Shed Three-Month
Sales Moving Average
January 10
February 12
March 13
April 16 (10+12+13)/3 = 11 2/3
May 19 (12+13+16)/3 = 13 2/3
June 23 (13+16+19)/3 = 16
July 26 (16+19+23)/3 = 19 1/3
11
Weighted Moving Averages
Weighted moving averages use weights to put more
emphasis on recent periods.
Weighted moving average =
12
Weighted Moving Averages
Weights Applied
Period
3 Last month
2 Two months ago
1 Three months ago
3*last month demand+2* two months ago demand+1*three months ago
demand
6 Sum of weights
13
Weighted Three-Month Moving Average
Month Actual Three-Month Weighted
Sales Moving Average
January 10
February 12
March 13
April 16 [3*13+2*12+1*10]/6 = 12 1/6
May 19 [3*16+2*13+1*12]/6 =14 1/3
June 23 [3*19+2*16+1*13]/6 = 17
July 26 [3*23+2*19+1*16]/6 = 20 1/2
14
Exponential Smoothing
Exponential smoothing is a type of moving average
technique that involves little record keeping of past
data.
New forecast
= previous forecast + (previous actual –previous forecast)
Mathematically this is expressed as:
Ft = Ft-1 + (Dt-1 - Ft-1)
Ft = new forecast
Ft-1 = previous forecast
Dt-1 = previous period actual
= smoothing constant (0< <1)
Exponential Smoothing
Qtr Actual Rounded Forecast using =0.10
1 180 175
2 168 176= 175.00+0.10(180-175)
3 159 175 =175.50+0.10(168-175.50)
4 175 173 =174.75+0.10(159-174.75)
5 190 173 =173.18+0.10(175-173.18)
6 205 175 =173.36+0.10(190-173.36)
7 180 178 =175.02+0.10(205-175.02)
8 182 178 =178.02+0.10(180-178.02)
9 ? 179= 178.22+0.10(182-178.22)
Exponential Smoothing
Qtr Actual Tonnage Rounded Forecast using =0.50
Unloaded
1 180 175
2 168 178 =175.00+0.50(180-175)
3 159 173 =177.50+0.50(168-177.50)
4 175 166 =172.75+0.50(159-172.75)
5 190 170 =165.88+0.50(175-165.88)
6 205 180 =170.44+0.50(190-170.44)
7 180 193 =180.22+0.50(205-180.22)
8 182 186 =192.61+0.50(180-192.61)
9 ? 184 =186.30+0.50(182-186.30)
Exponential Smoothing with Trend
Adjustment( Holt’s model)
Simple exponential smoothing fails to respond to trends, so
a more complex model is necessary with trend adjustment.
Simple exponential smoothing - first-order
smoothing
Trend adjusted smoothing - second-order
smoothing
Low gives less weight to more recent trends,
while high gives higher weight to more recent
trends.
Exponential Smoothing with Trend Adjustment
Forecast including trend (Ft+1)
= new forecast (Ft) + trend correction(Tt)
Ft = Dt-1 +(1- )(Ft-1 + Tt-1)
Tt = (1 - )Tt-1 + (Ft – Ft-1)
where
Tt = smoothed trend for period t
Tt-1 = smoothed trend for the preceding period
= trend smoothing constant
Ft = simple exponential smoothed forecast for period t
Ft-1 = forecast for period t-1
Example: Compute the adjusted exponential forecast for
the first week of march for a firm with the following data.
Assume the forecast for the first week of January (F0) as
600 and the corresponding initial trend (T0) as 0. let =
0.1 and =0.2.
Month
Jan. Feb.
Week 1 2 3 4 1 2 3 4
Demand 650 600 550 650 625 675 700 710
20
Solution: first week of jan.
Ft = Dt-1 +(1- )(Ft-1 + Tt-1) = 0.1 (650) + 0.9 (600 +0) = 605
Tt = (Ft – Ft-1)+ (1 - )Tt-1 = 0.2(605 - 600)+0.8(0)=1.00
Ft+1 = Ft + Tt = 605+1=606, so for first week of march is 644.04,
i.e 644 units.
Week Ft-1 Dt-1 Ft Tt Ft+1
jan 1 600.00 650 605.00 1.00 606.00
2 605.00 600 605.40 0.880 606.38
3 605.40 550 600.65 -0.246 600.40
4 600.65 650 605.36 0.742 606.10
Feb. 1 605.36 625 607.99 1.120 609.11
2 607.99 675 615.70 2.440 618.14
3 615.70 700 626.33 4.080 630.41
4 626.33 710 738.37 5.670 644.04
21
Trend- and Seasonality-Corrected Exponential
Smoothing (Winter’s Model)
• Appropriate when the systematic component of
demand is assumed to have a level, trend, and seasonal
factor
• Systematic component = (level+trend)(seasonal factor)
• Assume periodicity of demand to be p.
• Obtain initial estimates of level (L0), trend (T0), seasonal
factors (S1,…,Sp) using procedure for static forecasting
• In period t, the forecast for future periods is given by:
Ft+1 = (Lt+Tt)(St+1) and Ft+n = (Lt + nTt)St+n
Trend- and Seasonality-Corrected
Exponential Smoothing (continued)
After observing demand for period t+1, revise estimates for level,
trend, and seasonal factors as follows:
Lt+1 = (Dt+1/St+1) + (1-)(Lt+Tt)
Tt+1 = (Lt+1 - Lt) + (1-)Tt
St+p+1 = g(Dt+1/Lt+1) + (1-g)St+1
= smoothing constant for level
= smoothing constant for trend
g = smoothing constant for seasonal factor
Regression Analysis
In a simple regression analysis the relationship
between the dependent variable y and some
independent variable x can be represented by a
straight line
y= a+bx
Where, b is the slope of the line
a is the y-intercept
∑y= Na +b∑x , (i)
∑xy= a∑x + b∑x2 , (ii)
24
Example: the following data gives the sales of
the company for various years. Fit the straight
line. Forecast the sales for the year 2020.
year 2011 2012 2013 2014 2015 2016 2017 2018 2019
Sales 13 20 20 28 30 32 33 38 43
(000)
25
Year Sale (y) Deviation (x) x2 xy
1 13 -4 16 -52
2 20 -3 9 -60
3 20 -2 4 -40
4 28 -1 1 -28
5 30 0 0 0
6 32 1 1 32
7 33 2 4 66
8 38 3 3 114
9 43 4 16 172
N=9 ∑y= 257 ∑x=0 ∑x2 =60 ∑xy = 204
a = 28.56, b= 3.4
The equation of the straight line of best fit is
y= 28.56 + 3.4 x
So, sale for the year 2020 = 28.56 + 3.4 X 5 = 45.56= 45560
26
Forecasting Performance
How good is the forecast?
Forecast errors allow one to see how well the
forecast model works and compare that model with
other forecast models.
• Mean Forecast Error (MFE or Bias): Measures average deviation of forecast
from actuals.
• Mean Absolute Deviation (MAD): Measures average absolute deviation of
forecast from actuals.
• Mean Absolute Percentage Error (MAPE): Measures absolute error as a
percentage of the forecast.
• Standard Squared Error (MSE): Measures variance of forecast error
Measures of Forecast Accuracy
Forecast error = actual value – forecast value
Forecasting Performance Measures
1 n
MFE ( Dt Ft )
n t 1
1 n
MAD Dt Ft
n t 1
100 n Dt Ft
MAPE
n t 1 Dt
1 n
MSE ( Dt Ft ) 2
n t 1
Mean Forecast Error (MFE or Bias)
1 n
MFE ( Dt Ft )
n t 1
• Want MFE to be as close to zero as possible -- minimum bias
• A large positive (negative) MFE means that the forecast is
undershooting (overshooting) the actual observations
• Note that zero MFE does not imply that forecasts are perfect
(no error) -- only that mean is “on target”
• Also called forecast BIAS
Mean Absolute Percentage Error
(MAPE)
100 n Dt Ft
MAPE
n t 1 Dt
• Same as MAD, except ...
• Measures deviation as a percentage of actual
data
Mean Squared Error (MSE)
1 n
MSE ( Dt Ft ) 2
n t 1
• Measures squared forecast error -- error variance
• Recognizes that large errors are disproportionately
more “expensive” than small errors
• But is not as easily interpreted as MAD, MAPE -- not as
intuitive
Tracking signal
• Should be within the range of +6
• Otherwise, possibly use a new forecasting
method
TSt = bias / MADt
33
Hospital Days – Forecast Error Example
Month Forecast Actual
Ms. Smith forecasted
JAN 250 243
total hospital
FEB 320 315
inpatient days last
MAR 275 286
year. Now that the APR 260 256
actual data are known, MAY 250 241
she is reevaluating JUN 275 298
her forecasting JUL 300 292
model. Compute the AUG 325 333
MAD, MSE, and SEP 320 326
MAPE for her OCT 350 378
NOV 365 382
forecast.
DEC 380 396
Forecast Error – Example
Forecast Actual |error| error^2 |error/actual|
JAN 250 243 7 49 0.03
FEB 320 315 5 25 0.02
MAR 275 286 11 121 0.04
APR 260 256 4 16 0.02
MAY 250 241 9 81 0.04
JUN 275 298 23 529 0.08
JUL 300 292 8 64 0.03
AUG 325 333 8 64 0.02
SEP 320 326 6 36 0.02
OCT 350 378 28 784 0.07
NOV 365 382 17 289 0.04
DEC 380 396 16 256 0.04
MAD = MSE = MAPE = .0368*100
11.83 192.83 = 3.68
Capacity Planning
36
Capacity
Productive Capacity, generally measured in physical
units, refers either to the maximum output rate for
products or services or to the amount of key
resources available in each operating period.
37
Strategic capacity planning
Capacity is the ability to hold, receive store or accommodate raw material,
finished products, customers, etc.
Strategic capacity planning is an approach for determining the overall
capacity level of capital intensive resources. Including facilities, equipment,
and overall labour force size.
Capacity used is the rate of output actually achieved.
The best operating level is nominally the capacity for which the process was
designed.
Capacity Utilization rate = Capacity used / best operating level
38
Capacity Decision
Capacity-
• Maximum capacity to produce
• Rated capacity is theoretical
• Effective capacity includes efficiency and utilization
Capacity utilization
-percent of available time spent working
Capacity efficiency
-how well a machine or worker performs compared to a standard output level
Capacity load
-standard hours of work assigned to a facility
Capacity load percentage
- Ratio of load to capacity
39
Dimension of demand Effect on capacity requirements
quantity How much capacity is needed?
Timing When should capacity be available?
quality What kind of capacity is needed?
location Where should capacity be installed
40
Capacity measurement
Types of business Input measures of output measures
capacity of capacity
Car manufacturer Labour hours Cars per shift
Hospital Available beds Patients per
month
Pizza point Labour hours Pizzas per day
Retail store Floor space Revenue
41
42
Best/optimum operating level
The best operating level is the output that
results in the lowest average unit cost.
Economics of scale:
- Where the cost per unit of output drops as
volume of output increases
- Spread the fixed costs of buildings &
Equipments over multiple units, allow
bulk purchasing & handling of material
43
44
45
Making capacity planning decisions
The three-step procedure for making capacity planning decisions
is as follows:
Step1: Identify capacity requirements
Step2: Develop capacity alternatives
Step3: Evaluate capacity alternatives
Determine project
capacity requirements Formulate alternatives to meet future
given a demand capacity requirements
forecast
Evaluate alternatives based on cost, revenues, risks,
competition, flexibility, quality, organizational and managerial
adjustment
Select optimum alternatives and
implement capacity development plan 46
Decision Tree
A decision tree is a schematic model of the steps in the capacity
planning problem
Step:
1. Draw the various decisions
2. Add the possible states of nature, probabilities, and payoffs
3. Determine the expected value of each decision
4. Make decision (the one with maximum expected value)
47
Decision Tree Example
A glass factory specializing in crystal experiences substantial backlog and
management is considering three courses of action:
A. Arrange for subcontracting
B. Construct new facilities
C. Do nothing (no change)
The correct choice depends largely upon demand which may be low, medium, or
high. By consensus management estimates the respective demand probabilities as
0.1, 0.5, 0.4.
The management also estimates the profits when choosing from the three
alternatives under the differing probable levels of demand these profits are as
follows.
Low (p=0.1) Medium (p=0.5) High (p=0.4)
A 10000 50000 90000
B -120000 25000 200000
48
C 20000 40000 60000
EVA = 0.4*90 + 0.5*50 + 0.1*10 = 62 K H (0.4) 90 K
M (0.5)
A 50 K
L (0.1)
10 K
A
EVB = 80.5 K H (0.4)
200 K
M (0.5)
B 25 K
Start B L (0.1)
-120 K
C H (0.4) 60 K
EVc = 46 K M (0.5) 40 K
C
L (0.1)
20 K
Choose decision B highest expected value
49
Thank you
50