Professional Documents
Culture Documents
Week 2-Demand Management
Week 2-Demand Management
2023
Chapter 2
(Vollmann, Berry, Whybark & Jacobs, 2005)
Agenda
Demand management in MPC systems
Demand management in different manufacturing
environments
Communicating with other MPC modules and
customers
Information use in Demand Management
Day-to-day management activities required to
manage demand.
1
10.10.2023
Demand Management
Covers issues concerned with how a firm
integrates information from and about its
customers demand into the manufacturing
planning and control systems.
These customers are either internal or external
to the firm.
Demand Management
It includes several activities such as:
Determining & estimating (forecasting)
customer demand
Entering orders,
Determining specific product requirements
Converting customer orders into promised
delivery dates (due date promising)
Helping balance demand with supply.
2
10.10.2023
Demand Management
Proper planning of internal & external demand,
means capacity can be better managed &
controlled.
Timely & honest delivery dates can be possible.
Physical distribution activities can be improved
significantly.
3
10.10.2023
Forecasts
& Orders
SOP develops sales and operations plans for a year or more
(aggregate prod. plan) based on forecasted demand.
Both forecast and actual demand information is provided to the
MPS (Master Production Scheduling) module.
In the MPS, short-term, product-specific manufacturing plans are
developed and controlled as actual demand becomes available, and
information for delivery promises are provided to the customers.
END 421E- Production Planning and Control 7
4
10.10.2023
5
10.10.2023
6
10.10.2023
Information Use in DM
CRM – Individual customer data is
collected by Customer Relationship
Management software.
In MTS firms, the customer information at this
level can help determine early demand and mix
trends.
In MTO/ATO, CRM can be used to develop
similar insights into customers. Data can be
used to develop make-to-knowledge plans on
an individual customer basis.
7
10.10.2023
Agenda
Framework for Forecasting
Long, intermediate and short term forecasting
Data Cleaning
Regression Analysis and Cyclic Decomposition
Techniques for intermediate-term forecasts
Decomposition of a Time Series
Short-term forecasting techniques:
Techniques useful for short-term individual item
forecasts.
The metrics of forecast quality (error)
8
10.10.2023
9
10.10.2023
10
10.10.2023
11
10.10.2023
12
10.10.2023
Forecasting Concepts
Data Cleaning
Regression Analysis and Cyclic
Decomposition Techniques
Moving Averages Forecasting
Exponential Smoothing Forecasting
Evaluating Forecasts (Bias & MAD)
Data Cleaning
Past data may include demand periods where
promotions, discounts or other activities or events that
affected the demand in a negative or positive way.
These periods’ demand data has extreme values
(outliers)!
So these data must be cleaned.
Otherwise, we can reach a biased/inaccurate forecast
model.
13
10.10.2023
Data Cleaning
Standard deviation of data (σ) should be computed!
Lower limit=Mean-1*σ
Upper limit=Mean+1*σ
Trend - Data Cleaning
1000000
900000
800000
700000
600000 Data
Alt Limit
500000
Üst Limit
400000 Trendline
Cleaned Data
300000
200000
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
14
10.10.2023
Regression Analysis
Regression Analysis is used to forecast the demand
that has an increasing or decreasing trend!
Regression can be defined as a functional relationship
between two or more correlated variables.
Linear regression: The relationship between variables
forms a straight line.
Y= a+bX
Y: Value of the dependent variable (Demand)
a: is the Y-intercept (a constant value)
b: The slope, and X: Independent variable.
(In time series analysis, X is units of time.)
END 421E- Production Planning and Control 34
Regression Analysis
Linear regression is useful for medium term forecasting
of major occurrences and aggregate planning.
Though demand for individual products within a family
may vary widely during a time period, demand for the
total product family is surprisingly smooth.
So, linear regression would be very useful to forecast
demands for product families.
The major restriction: Future forecasts are assumed to
fall about a straight line.
Regression analysis is used when demand has a trend!
15
10.10.2023
Regression Analysis
Least squares method is used to estimate a, b
parameters. Y=a+bX
Example: A firm’s sales for a product line during the 12
quarters of the past three years were as follows:
The firm wants to forecast each quarter of the fourth
year—that is, quarters 13, 14, 15, and 16
Regression Analysis-
Least Squares
The least squares method tries to fit the line to the
data that minimizes the sum of the squares of the
vertical distance between each data point and its
corresponding point on the line.
The difference between the point and the line is y - Y.
This difference is the forecast error .
The sum of the squares of the differences:
16
10.10.2023
Regression Analysis-
Least Squares
(1)
(2)
Regression Analysis-
Least Squares
17
10.10.2023
Regression Analysis-
Least Squares
The standard error of estimate Syx (how well the line fits
the data):
18
10.10.2023
Regression Example
Total
19
10.10.2023
Trend Component
Seasonal Component
(Annual)
Cyclic Component
(2-10 years)
Random Component
20
10.10.2023
Seasonality
When demand contains both seasonal and
trend effects at the same time, the question
is how they relate to each other.
Two types of seasonal variation: additive and
multiplicative.
21
10.10.2023
Seasonality
Additive seasonal variation simply assumes that the
seasonal amount is a constant no matter what the
trend or average amount is.
Forecast including trend and seasonal =Trend + Seasonal
Seasonality
Multiplicative Seasonal Variation
The seasonal variation increasing as the trend
increases because its size depends on the trend.
The multiplicative seasonal variation is the usual
experience.
Essentially, this says that the larger the basic amount
projected, the larger the variation around this that we
can expect.
22
10.10.2023
Seasonal Factor/Index
A seasonal factor is the amount of correction
needed in a time series to adjust for the season of
the year.
We usually associate seasonal with a period of the
year characterized by some particular activity.
The following examples show how seasonal
indexes are determined and used to forecast
(1) a simple calculation based on past seasonal data
and
(2) the trend and seasonal index from a hand-fit
regression line
END 421E- Production Planning and Control 50
23
10.10.2023
24
10.10.2023
55
25
10.10.2023
56
26
10.10.2023
Forecast
demand for
the four
quarters of
the fourth
year.
59
27
10.10.2023
28
10.10.2023
29
10.10.2023
Total
END 421E- Production Planning and Control 65
30
10.10.2023
Period
31
10.10.2023
32
10.10.2023
33
10.10.2023
i = period number
t = current period (the period for which the most recent actual
demand is known)
n = number of periods in the moving average
END 421E- Production Planning and Control 72
Example
for period 33
for period 33
34
10.10.2023
35
10.10.2023
Exponential Smoothing-Example
ESFt=*Actual demandt-1+ (1- )*ESFt-1
ESF27=1000 (given!)
α=0.1
ESF28=0.1*900+0.9*1000=990
ESF29=0.1*1100+0.9*990=1001
For more volatile demand patterns, a bigger α (α>0.50)
should be preferred! Larger values of α give more weight
to recent demands and utilize older demand data less.
END 421E- Production Planning and Control 77
36
10.10.2023
Comparison
Comparison
ESF26 = 0.3 (Period 26 actual demand) + 0.7 (ESF25)
ESF27 = 0.3 (Period 27 actual demand)+ 0.7 [0.3 (Period
26 actual demand) + 0.7 (ESF25)]
Weight of 0.21 being applied to the actual demand in
period26 .
37
10.10.2023
Evaluating Forecasts
In order to evaluate each forecasting technique’s
forecasting performance individually;
OR
To compare different forecasting techniques’ to decide
the most suitable one for our data, we must use
measures of forecast error.
The technique with the least forecasting error is the
best.
38
10.10.2023
Evaluating Forecasts
Mean Error (bias): Dt: Actual (Past) Demand
n
(Dt − X t )
1 Xt : Forecasted Demand
ME = n: Number of Periods considered for
n t =1 forecast.
Evaluating Forecasts
Mean absolute deviation (MAD):
1 n
MAD = Dt − X t
Dt: Actual (Past) Demand
Xt : Forecasted Demand
n t =1 n: Number of Periods considered for
forecast.
39
10.10.2023
Evaluating Forecasts
Evaluating Forecasts
Mean Square Error (MSE):
1 n
MSE = (Dt − X t )
2 Dt: Actual (Past) Demand
Xt : Forecasted Demand
n t =1 n: Number of Periods considered
for forecast.
Advantage: All positive & negative error terms are considered in erro
computation.
Disadvantage: It does not clearly Show how precisely you forecasted, i.e. the
% of error.
40
10.10.2023
Concluding Principles
DM systems and procedures must be aligned
with the market environment of the firm
Data capture must not be limited to sales but
should include domain info such as
knowledge, trends, systems performance
Forecasting models should not be more
complicated than necessary. Simple
models work just as good.
41
10.10.2023
Concluding Principles
Input data and output forecasts should be
routinely monitored for quality and
appropriateness.
Information on sources of variation
should be incorporated into the
forecasting system.
Forecast from different sources must be
reconciled and made consistent with firm
plans and constraints.
42