Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
in
Supply
Chains
Dr Niniet
Indah
Arvitrida
Graduate Program
Department
of Industrial
Engineering
Institut
Teknologi
Sepuluh
Nopember
2019
1
In
this
course
students
will be
able
to
1. Understand
the
role
of
forecasting
for
both
an
enterprise
and
a
supply
chain
2. Identify
the
components
of
a
demand
forecast
3. Forecast
demand
using
time-‐series
methodologies
4. Analyze demand
forecast
error
Main source:
2
What is Forecasting ?
Y
Forecasting
4
Role of Forecasting
in a Supply Chain
The basis for all planning decisions in a supply
chain
Used for both push and pull processes
Production scheduling, inventory, aggregate planning
Sales force allocation, promotions, new production
introduction
Plant/equipment investment, budgetary planning
Workforce planning, hiring, layoffs
All of these decisions are interrelated
Characteristics of forecasting
Forecasts are usually wrong/inaccurate, seldom
correct, and involves error
à find the best method
Long-term forecasts are usually less accurate than short-
term forecasts.
Family / grouped / aggregate forecasts are usually more
accurate than item forecast
In general, the further up the supply chain a company is (the
further they are from the customers), the greater the
distortion of information they receive.
Forecast should include a measure of forecast error
6
Components and Methods
Companies must identify the factors that influence
future demand and then ascertain the relationship
between these factors and future demand
Past demand
Lead time of product replenishment
Planned advertising or marketing efforts
Planned price discounts
State of the economy
Actions that competitors have taken
Components and Methods
1. Qualitative
Primarily subjective
Rely on judgment
2. Time Series
Use historical demand only
Best with stable demand
3. Causal
Relationship between demand and some other factor
4. Simulation
Imitate consumer choices that give rise to demand
Assumptions of Forecast
9
Which
item
should
we
forecast?
Independent Demand: Dependent Demand:
Finished Goods;; spare parts Raw Materials,
Component parts,
Sub-assemblies, etc.
B(4) C(2)
10
Production System
Customer
order
decoupling
point
ETO
MTO
ATO
MTS
14
Forecasting
Steps
• Decide
what
needs
to
be
forecast
• Level
of
detail,
units
of
analysis (weekly,
monthly,
etc) &
time
horizon
required
• Evaluate
and
analyze
appropriate
data
• Identify
needed
data
&
whether
it’s
available
• Select
and
test
the
forecasting
model
• Cost,
ease
of
use
&
accuracy
• Generate
the
forecast
• Monitor
forecast
accuracy
over
time
15
Basic
approach
to
demand
forecasting
16
Characteristics
of
a
Good
forecast
• Accuracy
à bias
(when
forecast
are
persistently
too
high
or
too
low)
&
consistency
(concerned
with
the
size
of
forecast
error)
• Cost
• Response à stability
&
responsive
• Simplicity à easy
to
make,
understand
&
use
17
Measuring
Accuracy,
Forecast
Errors
To
compare
different
time
series
techniques
or
to
select
the
“best”
set
of
initial
values
for
the
parameters,
use
a
combination
of
the
the
following
3
metrics (assumption:
errors
are
normally
distributed):
n
• Mean
Absolute
Deviation (MAD)
Most
popular å A
i =1
i - Fi
Standard
deviation
of
errors
is
assumed MAD =
by
1.25
times
MAD
n
Assumption:
error
is
normally
distributed
n
• Mean
Square
Error (MSE) (
å i i
A - F )2
Note for this slide: E is error;; F is forecasted demand;; D is actual demand
19
• Forecast
error
measurements:
If
you
are
working
with
a
low-‐volume
or
sporadic
item
then
MAD
is
a
good
choice,
while
the
MAPE
should
be
avoided.
MAPE
is
good
for
high
volume
/
fairly
consistent
and
regular
demand
pattern.
20
Examples
of
Production
Resource
Forecasts
Forecast Units of
Time Span Item Being Forecast
Horizon Measure
• Product lines
• Factory capacities
• Planning for new products Dollars, tons,
Long-Range Years • Capital expenditures etc.
• Facility location or expansion
• R&D
• Product groups
• Department capacities
Medium- Dollars, tons,
Months • Sales planning
Range • Production planning and etc.
budgeting
• Specific product quantities
• Machine capacities
• Planning
• Purchasing Physical units
Short-Range Weeks • Scheduling of products
• Workforce levels
• Production levels
• Job assignments
21
Demand
Behavior
Trend
◦ a
gradual,
long-‐term
up
or
down
movement
of
demand
◦ Due
to
population,
technology
etc
Random
variations
◦ movements
in
demand
that
do
not
follow
a
pattern
Cycle
◦ an
up-‐and-‐down
repetitive
movement
in
demand
◦ Non-‐annual;
multi-‐year
◦ Due
to
interactions
of
factors
influencing
economy
Seasonal
pattern
◦ an
up-‐and-‐down
repetitive
movement
in
demand
occurring
periodically
◦ Occurs
within
one
year
◦ Due
to
weather,
customs
etc
22
Forms
of
Forecast
Movement
often regarded
as part of the
trend
Demand
Demand
Random
movement
Time Time
(a) Trend (b) Cycle
Demand
Demand
Time Time
(c) Seasonal pattern (d) Trend with seasonal pattern
23
Types
of
Forecasts
Qualitative
(Extrinsic) :
◦ requiring
no
manipulation
of
data,
◦ used
only
judgment of
the
forecaster,
◦ Evaluating
factors
other
than
historical
data
◦ concern
with
identifying
various
factors
that
can
influence
demand
◦ suitable
for
long
term forecasting
Quantitative
(Intrinsic) :
◦ no input
of
judgment,
◦ mechanical
procedure that
produce
quantitative
result,
◦ Rely
on
historical
information
◦ using
statistical
techniques
◦ suitable
for
short
term forecasting
24
1.
Qualitative
Methods
(Extrinsic)
25
Qualitative
Methods :
Example
Type Characteristics Strengths Weaknesses
Executive A group of managers Good for strategic or One person's opinion
opinion meet & come up with new-product can dominate the
a forecast forecasting forecast
Market Uses surveys & Good determinant of It can be difficult to
research interviews to identify customer preferences develop a good
customer preferences questionnaire
26
1.
Qualitative
Methods
-‐in
general
-‐
Drawbacks :
Advantages :
– High cost à for paying
– No historical data is labour time of the
required. experts
– Involving experts for – Time consuming,
modeling, so the results are particularly in building
quite accurate. the models and selecting
appropriate methods for
validation
27
2.
Quantitative
methods
(Intrinsic)
• Additive
:
Systematic
component
=
level
+
trend
+
seasonal
factor
• Mixed
:
Systematic
component
=
(
level
+ trend
)
x
seasonal
factor
30
Time
Series
Models
ORDERS
FORECAST
MONTH PER
MONTH
Jan 120 -‐
Feb 90 120
Mar 100 90
Apr 75 100
May 110 75
June 50 110
July 75 50
Aug 130 75
Sept 110 130
Oct 90 110
Nov
-‐ 90
33
Simple
Moving
Average
n
S
i = 1
Di
MAn =
n
where
34
Example
:
MA(3)
ORDERS MOVING
3
MONTH PER
MONTH AVERAGE S
i =
1
Di
Jan 120 – MA3 =
Feb 90 – 3
Mar 100 –
Apr 75 103.3 90
+
110
+
130
= 3
May 110 88.3
June 50 95.0
July 75 78.3 =
110
orders
Aug 130 78.3
for
Nov
Sept 110 85.0
Oct 90 105.0
Nov -‐ 110.0
35
Example
:
MA(5)
ORDERS MOVING
5
MONTH PER
MONTH AVERAGE
Jan 120 –
S
i =
1
Di
Feb 90 – MA5 =
Mar 100 –
5
Apr 75 –
90
+
110
+
130+75+50
May 110 – =
June 50 99.0
5
July 75 85.0
Aug 130 82.0 =
91
orders
Sept 110 88.0 for
Nov
Oct 90 95.0
Nov -‐ 91.0
36
Smoothing
Effects
150 –
125 – 5-month
100 –
Orders
75 –
50 – 3-month
Actual
25 –
0 – | | | | | | | | | | |
Jan Feb Mar Apr May June July Aug Sept Oct Nov
Month
37
2.
Weighted
Moving
Average
WMAn = S Wi Di
i = 1
where
Wi = the weight for period i,
between 0 and 100
percent
S Wi = 1.00
38
Weighted
Moving
Average
Example
= 103.4 orders
39
Disadvantages of M.A. Methods
Ft = a Dt-1 + (1 - a) a Dt-1 + (1 - a)2 a Ft-1
And soon
42
Exponential
Smoothing:
Concept
• Include
all
past
observations
• Weight
recent
observations
much
more
heavily
than
very
old
observations:
0 < a < 1
weight
Decreasing weight given a
to older observations
a (1 - a )
a (1 - a ) 2
a (1 - a ) 3
today
!
43
Exponential
Smoothing
(α=0.30)
Ft =
a Dt-1 +
(1
-‐ a)Ft-1
PERIOD MONTH DEMAND F2 = aD1 + (1 - a)F1
1 Jan 37 = (0.30)(37) + (0.70)(37)
2 Feb 40 = 37
3 Mar 41
4 Apr 37 F3 = aD2 + (1 - a)F2
5 May 45 = (0.30)(40) + (0.70)(37)
6 Jun 50
= 37.9
7 Jul 43
8 Aug 47
F13 = aD12 + (1 - a)F12
9 Sep 56
10 Oct 52 = (0.30)(54) + (0.70)(50.84)
11 Nov 55 = 51.79
12 Dec 54
44
Exponential
Smoothing
(cont.)
FORECAST, Ft + 1
PERIOD MONTH DEMAND (a = 0.3) (a = 0.5)
1 Jan 37 37 37
2 Feb 40 37.00 37.00
3 Mar 41 37.90 38.50
4 Apr 37 38.83 39.75
5 May 45 38.28 38.37
6 Jun 50 40.29 41.68
7 Jul 43 43.20 45.84
8 Aug 47 43.14 44.42
9 Sep 56 44.30 45.71
10 Oct 52 47.81 50.85
11 Nov 55 49.06 51.42
12 Dec 54 50.84 53.21
13 Jan – 51.79 53.61
45
Exponential
Smoothing
(cont.)
70 –
60 – Actual a = 0.50
50 –
40 –
Orders
a = 0.30
30 –
Larger a, more responsive forecast;; Smaller a,
20 – smoother forecast
“Best” a can be found by Solver
10 – Suitable for relatively stable time series
0 – | | | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12 13
Month
46
Similarities
between MA
and ES
47
• What
happens
when
there
is
a
definite
trend?
560
550
Actual
540
Demand 530
Forecast
520
510
500
490
480
1 2 3 4 5Month
6 7 8 9 10
48
4.
Holt’s
– Trend
Corrected
(double
exponential
smoothing)
Level
à St+1 =
aDt+1 +
(1-‐a)(St +
Gt)
Trendà Gt+1 =
b(St+1 – St)
+
(1-‐b)Gt
D = Actual Demand
S = Intercept (level)
G = Slope (trend)
• More
stability
is
given
to
the
slope
estimate
which
implies
β ≤
α.
• τ is
the
amount
of
step
ahead
forecast
49
The
larger
the
Beta
value
leads
to
a
more
progressive
forecasting
results,
in
terms
of
modeling
trend.
350
300
250 Demand
Holt B(0.1)
200 Holt B(0.3)
150
100
1 2 3 4 5 6 7 8 9 10 11 12
50
Holt’s
– Trend
Corrected
• alfa
=
0,1
;
beta
=
0,1
52
Holt’s
– Trend
Corrected
53
Holt’s
– Trend
Corrected
54
5.
Accommodating
Seasonality (No
Trend)
w1 w2 w3 W4
55
300
250
200
150
100
50
0
M T W T F M T W T F M T W T F M T W T F
56
Procedures
57
From
the
Above
Data
• Average
factors
F 1.37 1.43 1.33 1.48 1.40
corresponding
to
the
same
period
of
the
season
58
Forecasts
for
Next
Week
59
6.
Winter’s
– Trend
and
Seasonality
Corrected
or
60
Winter’s
-‐-‐ Static
Methods
where
63
Winter’s
– Trend
and
Seasonality
Corrected
The
three
steps:
64
Winter’s
– Trend
and
Seasonality
Corrected
p is the number of periods after which the seasonal cycle repeats
65
Deseasonalized
data
(for
Winter’s)
Year Quarter Black Plastic Deseasonalized
N = 20, p = 4, so use the “even” equation
1998 I 2250
II 1737
III 2412 3575
IV 7269 3783,75
D
1999 I 3514 3965,375
II 2143 4069,625
ingat,
bahwa
III 3459 4118,75
ada
5
cycle
IV 7056 4272,375 kejadian
2000 I 4120 4237,375 seasonal
(lihat:
1 II 2766 4274,125 sheet
Raw
III 2556 4595,125 Data)
IV 8253 4968,5
2001 I [2250
5491 +
3514
+
5390,375
(2*(1737+2412+7269))]/(2*4)
II 4382 6083
III 4315 6575,375
IV 12035 6509,25
2002 I 5648 6489,5
II 3696 6688,25
III 4843
IV 13097 66
67
Deseasonalized
data
(for
Winter’s)
Deseasonalized Pooled Seasonal
Year Quarter Black Plastic Deseasonalized Seasonal Factor
Regression Factor
D
1999 I 3514 3965,375 3727,027022 0,942842641
II 2143 4069,625 3953,882904 0,541998853
III 3459 4118,75 4180,738787 0,82736573
IV 7056 4272,375 4407,594669 1,600873159
2000 I 4120 4237,375 4634,450551 0,888994273
II 2766 4274,125 4861,306434 0,568982852
III 2556 4595,125
3 5088,162316 0,502342465
IV 8253 4968,5 5315,018199 1,552769848
Y = A + B*t
2001 I 5491 5390,375 5541,874081 0,990820058
II 4382 6083
Y t = 2592,75 + (226,85) 1
5768,729963 0,759612606
III 4315 6575,375 (where t = the sequence number of
5995,585846 0,719696142
IV 12035 6509,25 data ;; A = intercept ;; B = slope)
6222,441728 1,934128197
2002 I 5648 6489,5 6449,29761 0,875754282
II 3696 6688,25 6676,153493 0,553612197
III 2
4843 6903,009375 0,701578071
IV 13097 7129,865257 1,83692111
intercept 2592,74761
slope 226,8558824 68
Deseasonalized
data
(for
Winter’s)
Deseasonalized Pooled Seasonal
Year Quarter Black Plastic Deseasonalized Seasonal Factor
Regression Factor
These
constants
are from
the
We
obtain
the
initial
estimates
of
level
and
trend
exactly
previous
as
in
the
static
case
when
we
deseasonalized the
step
demand (deseaso
(S0 =
intercept,
G0 =
slope) nalized
demand-
step)
Note: the Winter’s method can be started from this step, if S0, G0, c1, c2, c3,
and c4 are known in advance 70
Winter’s
– Trend
and
Seasonality
Corrected
(using
Seasonal
Factor)
These
constants
are from
the
previous
St+1 = a(Dt+1/ct+1) + (1-a)(St+Gt) step
S1 =
(0,1)(2250/0,899) +
(deseaso
(0,9)(2592.75+226.86) nalized
demand-
step)
71
Winter’s
– Trend
and
Seasonality
Corrected
(using
Seasonal
Factor)
72
Winter’s
– Trend
and
Seasonality
Corrected
(using
Seasonal
Factor)
74
The
Role
of
IT
in
Forecasting