Specialty Packaging Corporation produces plastic containers made from polystyrene. Julie Williams wants to forecast demand for black and clear plastic containers from 2007-2009 to improve supply chain performance. Historical demand shows seasonality. Analysis identifies time series methods like CMA and seasonal indices are appropriate. Regression forecasts are made and seasonal adjustments applied. Coordination mechanisms like information sharing and alignment are recommended to reduce supply chain instability from unpredictable demand.
2. About Polystyrene
Polystyrene (PS) is a synthetic aromatic polymer made from the monomer styrene, a liquid
petrochemical. Polystyrene can be rigid or foamed. General purpose polystyrene is clear, hard
and brittle. It is a very inexpensive resin per unit weight. It is a rather poor barrier to oxygen
and water vapor and has a relatively low melting point. Polystyrene is one of the most widely
used plastics, the scale of its production being several billion kilograms per year. Polystyrene
can be naturally transparent, but can be colored with colorants. Uses include protective
packaging (such as packing peanuts and CD and DVD cases), containers (such as "clamshells"),
lids, bottles, trays, tumblers, and disposable cutlery.
As a thermoplastic polymer, polystyrene is in a solid (glassy) state at room temperature but
flows if heated above about 100 °C, its glass transition temperature. It becomes rigid again
when cooled. This temperature behavior is exploited for extrusion, and also
for molding and vacuum forming, since it can be cast into molds with fine detail.
3. Problem Identification
Julie Williams wants to :
Select the appropriate forecasting method and estimate
the likely forecast error. Which should she choose?
Forecast quarterly demand for each of the two types of
containers for the years 2007 to 2009.
Improve supply chain performance, as SPC had been
unable to meet demand effective over the previous
several years.
4. Supporting Theory
Forecasting Classifield
Qualitatif
Primarily subjective and rely on human judgment.
Causal
The demand forecast is highly corelated with certain factors
in the environment
Simulations
Imitate the consumer choices that give rise to demand to
arrive at a forecast
Time Series
Use historical demand to make a forecast
Multiplicate : level x trend x seasonal factors
Additive : level + trend + seasonal factors
Mixed : (level trend) x seasonal factors
Forecast Method Applicability
Moving average No trend or
seasonality
Simple exponential
smoothing
No trend or
seasonality
Holt’s model Trend but no
seasonality
Winter ‘s model Trend and
seasonality
5. Supporting Theory
Basic Approach to Demand Forecasting
Understand the objective
of forecasting
Integrate demand
planning and forecasting
Understand and identify
customer segment
Identify the major factors
that influence the
demand forecast
Determine the
appropriate forecasting
technique
Establish performance
and error measure for the
forecast
7. Analysis 1
Over the several years, they had
been unable to meet demand
Understand the objective
of forecasting
Integrate demand
planning and forecasting
Establish a collaborative forecast using
data from the SPC and Customer
Have two produts, black and clear plastic
Have quarterly historical demand plastic
container
9. Analysis 1
Increasing volume (‘000 lb) in
every quarter each years.
Historical demand of plastic
containers influence by seasonal
demand
Determine the
appropriate forecasting
technique
Establish performance
and error measure for the
forecast
MSA (Mean Square Error)
MAPE (Mean Absolute Percentage Error)
10. Analysis 2
Year Quarter
Black Plastic
Demand
Clear Plastic
Demand
2007
I 6,759 5,929
II 5,154 15,158
III 5,366 8,149
IV 13,864 4,190
2008
I 7,620 6,488
II 5,790 16,555
III 6,009 8,883
IV 15,476 4,559
2009
I 8,481 7,048
II 6,426 17,951
III 6,651 9,617
IV 17,087 4,928
REGRESI LINIER
Black Y = 8,886.63 + 853.79x
Clear Y = 15,001.69 + 700.61x
Year Quarter Sumbu X
Black plastic
demand
CMA
SEASONAL
RATIO
INDEX
DESEASONALI
ZED SALES =
Sales/Season
al Index
(Sumbu Y)
TREND (Y =
8,886.63 +
853.79x)
TREND AFTER
ADJUSTMENT
BY THE
SEASONAL
INDEX =
Trend x
Seasonal
Index
WEIGHT
2002 I 1 2,250 0.5 0.25 8,926.81 9,740.42 2,455
II 2 1,737 1 0.19 9,325.34 10,594.21 1,973
III 3 2,412 1 12,982.25 0.19 0.19 12,821.17 11,448.00 2,154
IV 4 7,269 1 14,331.38 0.51 0.47 15,402.99 12,301.79 5,805
Time series
CMA
Seasonal and trend
Ekstrapolasi regresi linier
Forecasting method
11. Analysis 3
Julie Williams used optimum forecast to meet
unpredictable demand influence by seasonal demand
(response supply chain objective)
Orders
0
Time
Sales from
store
Orders
0
Time
Store’s orders
to wholesaler
Manufacturer’s
orders to its
suppliers
Orders
0
Time
Wholesaler’s
orders to
manufacturer
Orders
0
Time
Retail
Store
Whole
-saler
Manuf-
acturer
Supplier
12. Analysis 3
Coordination
mechanism for reducing
supply chain dynamic
instability by using
information sharing,
channel alingment and
operational efficiency
14. Lesson learned
Company should understand the role of forecasting for
both an enterprice and a supply chain.
Manage unpredictable demand with coordination
mechanism by using information sharing, channel
alingment and operational efficiency.