Specialty packaging corporation case study answers

Technology And Information Management

Problem 1: Planning

1. Define the Real Problem a. Look over all the questions below and create a plan that will help in completing this

exam successfully

• Make a list of the tasks that need to be done in order to complete this

examination successfully.

• Use GANTT charts to create a schedule for these tasks and keep track of them

accordingly.

b. When done, draw conclusions and develop guidelines to better your own strategies and implementation in the future

2. Plan a. What information is available for solving the problem?

• Lecture notes

o Notes on Activity Matrix

o Notes on GANTT chart

o Notes on PERT chart

• Canvas Handouts

b. What assumptions need to be made to make the solution process manageable? • Problem Solver

o Project Planner

• Audience:

o Professor Desa and TA’s

c. What analysis needs to be performed to resolve the issues defined in Step 1?

• Create a project schedule in order to allocate time successfully for this problem

o Determine all that is needing to be done on a high level for each

problem

o Create am activity matrix

o Create a GANTT chart

o Identify the “critical path” using a PERT Chart

o Keep track of each task and document completion

3. Execute the Plan a. Create a project schedule in order to allocate time successfully for this problem

Task Time Allocated

• Determine all that is needing to be done on a high level for each problem

15 minutes

• Create am activity matrix 20 minutes

• Create a GANTT chart Undetermined/Continuous

• Identify the “critical path” using a PERT Chart 15 minutes

• Keep track of each task and document completion Undetermined/Continuous

• Problem 2: Supply Chain Strategy for SPC

• Problem 3: Demand Forecasting for SPC

• Problem 4: Cycle Inventory for Polystyrene at SPC

• Problem 5: Safety Inventory for Polystyrene Resin at SPC

• Problem 6: Execution of Your Plan

b. Determine all that is needing to be done on a high level for each problem

Problem 1: Planning

(A) Create am activity matrix

(B) Create a GANTT chart

(C) Identify the “critical path” using a PERT Chart

(D) Keep track of each task and document completion

Problem 2: Supply Chain Strategy for SPC

(E) Read the Specialty Packaging Corporation Case Study

(F) Competitive Strategy

(G) Supply Chain Strategy

(H) Where does SPC lie in the zone of strategic fit between IDU and

responsiveness?

(I) Identify what SPC’s high-level SC strategy should be for each of the SC

drivers.

Problem 3: Demand Forecasting for SPC

(J) Form Hypothesis

(K) Forecast Demand for clear plastic using the 5 methods

(L) Identify the better method for Clear Plastic.

(M) Was the Hypothesis Correct?

(N) Forecast Demand for 2007 Clear Plastic.

Problem 4: Cycle Inventory for Polystyrene at SPC

(O) Why should SPC have a cycle inventory?

(P) What are the following values for clear plastic?

(Q) Short-Term Discounting

Problem 5: Safety Inventory for Polystyrene Resin at SPC

(R) Should SPC have a safety inventory? Why? How much safety inventory

would you recommend for SPC?

Problem 6: Execution of Your Plan

(S) Create table to compare your plan

c. Create am activity matrix

A B C D E F G H I J K L M N O P Q R S

A A x x

B B x

C x x C x

D x D

E E x

F x F x

G x x G x

H x x x H

I x x x x I x

J J x

K x K x

L x x L x

M x x x M x

N x x x x N x

O O

P x P

Q x x Q x

R x R

S x S

d. Create a GANTT chart

e. Identify the “critical path” using a PERT Chart

f. Keep track of each task and document completion Tasks Actual Time Allocated

A 20 Minutes

B 40 Minutes

C, D, E 15 Minutes Each

F, G, H, I 15 Minutes Each

J 5 Minutes

K 300 Minutes

L 10 Minutes

GANTT CHART PROJECT TITLE

PROJECT MANAGER

1 Problem 1: Planning

1.1 Activity Matrix Raqibul Mollah 2/7/19 2/7/19 1 100%

1.2 GANTT chart Raqibul Mollah 2/7/19 2/11/19 1 100%

1.3 PERT Chart Raqibul Mollah 2/7/19 2/11/19 1 100%

1.4

Keep track of each task and document

completion Raqibul Mollah 2/7/19 2/11/19 100%

2 Problem 2

2.1

Read the Specialty Packaging Corporation Case

Study Raqibul Mollah 2/8/19 2/8/19 1 100%

2.2 Competitive Strategy Raqibul Mollah 2/8/19 2/8/19 1 100%

2.3 Supply Chain Strategy Raqibul Mollah 2/8/19 2/8/19 1 100%

2.4

Where does SPC lie in the zone of strategic fit

between IDU and responsiveness? Raqibul Mollah 2/8/19 2/8/19 1 100%

2.5

Identify what SPC’s high-level SC strategy

should be for each of the SC drivers. Raqibul Mollah 2/8/19 2/8/19 1 100%

3 Problem 3

3.1 Form Hypothesis Raqibul Mollah 2/8/19 2/10/19 1 100%

3.2

Forecast Demand for clear plastic using the 5

methods Raqibul Mollah 2/8/19 2/10/19 1 100%

3.3 Identify the better method for Clear Plastic. Raqibul Mollah 2/8/19 2/10/19 1 100%

3.4 Was the Hypothesis Correct? Raqibul Mollah 2/8/19 2/10/19 1 100%

3.5 Forecast Demand for 2007 Clear Plastic. Raqibul Mollah 2/8/19 2/10/19 1 100%

4 Problem 4

4.1 Why should SPC have a cycle inventory? Raqibul Mollah 2/10/19 2/10/19 1 100%

4.2 What are the following values for clear plastic? Raqibul Mollah 2/10/19 2/10/19 1 100%

4.3 Short-Term Discounting 2/10/19 2/10/19 100%

5 Problem 5

5.1

Should SPC have a safety inventory? Why? How

much safety inventory would you recommend

for SPC? Raqibul Mollah 2/11/19 2/11/19 1 100%

6 Problem 6

6.1 Create table to compare your plan Raqibul Mollah 2/11/19 2/11/19 1 100%

PCT OF TASK

COMPLETE

THURSDAY FRIDAY SATURDAY SUNDAY MONDAYWBS

NUMBER TASK TITLE TASK OWNER

START

DATE END DATE DURATION

https://goo.gl/PXLbMe

TIM 125 Midterm COMPANY NAME TIM 125

Raqibul Mollah DATE 2/11/2019

A | 20 Min B | 40 Min C | 15 Min D | 15 Min E | 15 Min F | 15 Min

H | 15 Min I | 15 Min J | 5 Min K | 300 Min L | 10 Min M | 5 Min

G | 15 Min

N | 15 Min

O | 55 Min P | 20 Min Q | 15 Min R | 15 Min S | 30 Min

M 5 Minutes

N 15 Minutes

O 55 Minutes

P 20 Minutes

Q, R 15 Minutes

S 30 Minutes

4. Check your work a. Is the work correct in every detail?

• This work is correct in every detail because before even working on this

problem, intense planning was conducted to execute this problem. In order to

get every very accurate, the plan was created by following format from the

lecture notes and canvas handout.

b. Are my assumptions reasonable?

• My assumptions were reasonable because the assumptions were created from

intense planning and background knowledge.

c. In terms of the things I know, do the results make sense?

• In terms of the things I know, the results do make sense because intense

planning was conducted to execute this problem and these plans were created

by following format from the lecture notes.

5. Learn and Generalize a. I’m not a big fan of planning things through because it is a lot of work to properly plan

things out. But this exercise was great in the sense where it allowed me to stay on track

and not fall behind. Listing out all the things I needed to do also helped me allocate all

my time efficiently during finals week and not get stressed out. I will use this tactic when

planning out large assignment, so I can stay on track and produce high quality work in a

time efficient manner.

Problem 2: Supply Chain Strategy for SPC

1. Define the Real Problem a. What should SPC’s competitive strategy be?

b. What should SPC’s supply chain strategy be to align with its competitive strategy?

c. Where does SPC lie in the zone of strategic fit between IDU and responsiveness?

d. What should SPC’s high-level SC strategy be for each of the supply chain drivers?

2. Plan a. What information is available for solving the problem?

• Lecture Notes

• Textbook

• “Specialty Packaging Corporation” Case Study

b. What assumptions need to be made to make the solution process manageable?

• Problem Solver

o Market Strategist

• Audience:

o CEO or CIO of a company

c. What analysis needs to be performed to resolve the issues defined in Step 1?

• Read “Specialty Packaging Corporation” Case Study

• Identify competitive strategy

• Identify Supply chain strategy

• Place SPC in zone of strategic fit.

• Identify High Level SC strategy for each of the supply chain drivers.

3. Execute the Plan a. Competitive Strategy

• Porter’s Five (Six) Forces Analysis

Figure 2.1: Porter's Six Forces

Explanation of the six forces: • Competitors:

o The competitors force in this market is very high because there are

many companies that specialize in packaging considering plastic is quite

cheap. Most of these companies are privately owned thus do not have

exit strategies so many of them stay in the game since it is a steady

growing market.

• New Entrants: o The force for new entrants is low. Most new players are small scale

companies, and common technology to mold or create plastic packaging

is easily acquired. It’s hard to differentiate product when there are

multiple players producing plastic containers.

• Suppliers: o The force for suppliers is medium. There are several suppliers that

supply resin pellets and the machinery to convert the pellets into plastic

or thermoforming presses. Switching suppliers would not be a big deal

Competitors:

Amcor Berry Plastic

Tetra Pak

New Entrants:

Alpha Packaging

SKS

Buyers:

Supermarket

Consumers

Substitutes:

Ceraminc

Glass

Eco-Friendly Materials

Compliments:

N/A

Suppliers:

Polystyrene Resin supplier

depending on the costs and quality.

• Buyers: o The force for buyers is high. Consumers have a choice of whom to buy

from since there is no differentiation in producing plastic containers and

trays. There are many companies who product and replicate nearly the

same product thus customers hold some power.

• Substitutes: o The force for substitutes is low. Although there are many substitutes for

plastic containers and trays, the price of those substitutes are too

expensive to afford in huge amounts. Thus, the chance of those

materials replacing the volume of plastics is highly unlikely since one of

the main reasons plastic is so popular is because it is so inexpensive.

• Complements: o There are no complementors, maybe except plastic forks and spoon.

Some consumers might want to buy it from the same company at an

extent but not really.

• Berry Plastics focuses on a differentiated strategy because they provide

a variety of plastic containers ranging from household containers to

pharmaceutical containers. Amcor is well known around the world

therefore the focus on cost leadership to provide their customers with

high quality, low-price products. SPC’s competitive strategy runs with

Focus. SPC aims to provide its customers with recyclable/disposable

products through their polystyrene plastics.

Cost Leadership: Berry Plastic

Differentiation: Amcor

Focused: SPC

Source of Competitive Advantage

St ra

te gi

c Ta

rg e

t

Low Cost Differentiation

B ro

ad

N ar

ro w

Highly

Efficient

Somewhat

Efficient Somewhat

Responsive

Highly

Responsive

Low

Uncertainty

Somewhat Low

Uncertainty Somewhat High

Uncertainty

High

Uncertainty

b. Supply Chain Strategy

Figure 2.2: Supply Chain Stages

c. Where does SPC lie in the zone of strategic fit between IDU and

responsiveness?

Efficiency/Responsiveness Spectrum

• As a manufacturer and supplier of plastic containers, Specialty Packaging

Corporation should aim for a highly efficient supply chain to match with the

demand of the customers. By being highly efficient, SPC can maintain a low-cost

system. They are also a highly efficient supply chain because these plastics can

be made well in advance.

Implied Demand Uncertainty

• SPC’s products are considered a somewhat certain IDU. This is because they are

producing plastic packaging that is used daily. SPC isn’t at a low IDU level

because of the service levels that they are required to meet. Demand spikes for

Figure 2.3: Efficiency/Responsiveness Spectrum

Figure 2.4: Implied Demand Uncertainty

either clear plastic or black plastic depending on the season therefore SPC’s IDU

is somewhat certain.

• As seen in Figure 2.5 above, SPC lies in the lower section of IDU and

Responsiveness in the Zone of Strategic Fit. This is due to their product being

mature and thus relatively low demand uncertainty (note that a level of

uncertainty exists, due in part to the seasonality of demand and SPC’s past

ineffective demand forecasting). This in turn brings down responsiveness as SPC

is able to focus on making their supply chain more efficient.

d. What should SPC’s high-level SC strategy be for each of the supply chain

drivers?

• To develop a high-level supply chain strategy, we need to identify the six drivers

used to drive a supply chain. These are Facilities, Inventory, Transportation,

Information, Sourcing, and Pricing.

• Facilities: o These are actual physical locations in the SCN where products are

stored, assembled, or fabricated. There are two major types of facilities:

production sites, and storage sites. SPC needs to make the decision of

having facilities domestic or abroad. There is a trade-off between

transportation and production cost here. Outsourcing production to

countries like China may have lower production costs, but then

transportation cost to deliver the material back to the US will be higher.

• Inventory: o This encompasses all raw materials, work in process, and finished goods

within a supply chain. Inventory policies will determine SPC’s supply

chain efficiency and responsiveness. In SPC’s case, their Inventory

aspect of the supply chain is considered responsive because they store

High Responsiveness /

Low Efficiency

Low Responsiveness / High Efficiency

Low IDU High IDU

Figure 2.5: Zone of Strategic Fit

inventory of each type of sheet in anticipation for future demand.

• Transportation: o Transportation is moving inventory from point to point in the supply

chain. SPC can either focus on responsiveness by using a faster mode of

transportation for their products, however this will result in an increase

in transportation costs. The other option SPC has is to structure their

supply chain to provide next-day service using ground transportation.

This keeps responsiveness high with a low cost.

• Information: o This consists of data and analysis concerning facilities, inventory,

transportation, costs, prices, and customers throughout the supply

chain. Information helps improve a supply chain to be both responsive

and efficient. SPC can develop an IT infrastructure to manage customers

and their orders, and SPC’s suppliers as well. By matching supply with

demand, SPC can achieve a high level of responsiveness to customer

demand while keeping a lower production cost.

• Sourcing: o This is the choice of who will perform a supply chain activity. As I said

before in Facilities, SPC has the choice of deciding whether they want

lower production costs by sourcing their production in other countries.

They must decide whether they want efficiency or responsiveness.

• Pricing: o This determines how much SPC will charge for the goods and services it

provides. SPC should have differential because it provides

responsiveness to customers that value it and low cost to customers

that do not value responsiveness as much.

4. Check your work a. Is the work correct in every detail?

• After re-reading the “Specialty Packaging Corporation” case study, I can safely

assume that work I have written is correct.

b. Are my assumptions reasonable?

• My assumptions about the product and company are valid. Also, I have carefully

read and understood SPC’s manufacturing process to aid in my design of their

strategy for the supply chain and for each of the key drivers. All figures and

tables are labeled and explained, and all information is presented clearly.

c. In terms of the things I know, do the results make sense?

• Yes, in terms of the things I understood from my research, the results do make a

lot of sense.

5. Learn and Generalize a. A company must understand both its product and the market in order to successfully

design an effective supply chain. The supply chain strategy needs to complement the

competitive strategy in order for a supply chain to be successful. SPC’s functions within

their supply chain must be cohesive and work towards achieving the main goal. SPC

needs to focus more on matching customer demands because they did not have enough

product or inventory. As a product matures, the market demand is better understood

and thus the IDU goes down with time. This means the firm must design their supply

chain to be less responsive and more efficient to properly maximize the supply chain

profitability.

Problem 3: Demand Forecasting for SPC

1. Define the Real Problem a. What is your Hypothesis?

b. Which forecasting method should Julie Williams use for Clear Plastic?

c. Was your hypothesis correct?

d. What is the demand forecast for each quarter of 2007 for Clear plastic?

e. What is the annual demand for the year 2007?

2. Plan a. What information is available for solving the problem?

• “Specialty Packaging Corporation” Case Study

• Lecture Notes about Demand Forecasting

• Textbook

b. What assumptions need to be made to make the solution process manageable?

• Problem Solver

o Market Analyst

• Audience:

o CEO or CIO of a company

c. What analysis needs to be performed to resolve the issues defined in Step 1?

• Form a hypothesis

• Forecast Demand for clear plastic using the following methods:

o Static

o Moving Average

o Simple Exponential Smoothing

o Holt’s

o Winter’s

• Identify the better method for Clear Plastic.

• Was the Hypothesis Correct?

• Forecast Demand for 2007 Clear Plastic.

• Annual demand for the year 2007

3. Execute the Plan

Hypothesis:

a. I hypothesize that Julie Williams should use the Winter’s method as a forecasting

method for the clear plastic because from my understanding, Winter’s methods

uses all three of the smoothing constant, so it looks in multiple factors that

allows it to give a better forecasting for future demand.

Static Method a. Given Data

Figure 3.1: Given Demand Data for Clear Plastic

Year Period Quarter

Clear Plastic

Demand ('000 lbs)

1 I 3,200

2 II 7,658

3 III 4,420

4 IV 2,384

5 I 3,654

6 II 8,680

7 III 5,695

8 IV 1,953

9 I 4,742

10 II 13,673

11 III 6,640

12 IV 2,737

13 I 3,486

14 II 13,186

15 III 5,448

16 IV 3,485

17 I 7,728

18 II 16,591

19 III 8,236

20 IV 3,316

2002

2003

2004

2005

2006

b. Step 1: De-seasonalize demand in order to run linear regression to estimate

level and trend.

• Possible equations to use:

• We are measuring demand on a quarterly basis therefore, periodicity (p) is 4.

We start deseasonalizing demand at period (t) = 3.

• Excel formula used: =(D2+D6+2*SUM(D3:D5))/8

Figure 3.2: De-Seasonalized Demand for Clear Plastic

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

I 1 3,200

II 2 7,658

III 3 4,420 4,472

IV 4 2,384 4,657

I 5 3,654 4,944

II 6 8,680 5,049

III 7 5,695 5,132

IV 8 1,953 5,892

I 9 4,742 6,634

II 10 13,673 6,850

III 11 6,640 6,791

IV 12 2,737 6,573

I 13 3,486 6,363

II 14 13,186 6,308

III 15 5,448 6,932

IV 16 3,485 7,887

I 17 7,728 8,662

II 18 16,591 8,989

III 19 8,236

IV 20 3,316

2002

2003

2004

2005

2006

c. We then performed a regression analysis between the de-seasoned demand and

its period in order to determine level and trend.

• Level (L) is the intercept value 3,612 while T is the x variable coefficient of

263.94. We will round this up to make the equation for deseasonalized demand

= 3612 + 264t

• We obtain L = 3,612 and T = 264

• Regressed Equation: Regressed Demand = 3,612 + 264t

• Plugging =3,612 + C2*264 into Excel and copying down the “Regressed Data”

coulomb to apply the regression equation to all actual demand points resulted

in the following:

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

I 1 3,200 3876

II 2 7,658 4140

III 3 4,420 4,472 4404

IV 4 2,384 4,657 4668

I 5 3,654 4,944 4932

II 6 8,680 5,049 5196

III 7 5,695 5,132 5460

IV 8 1,953 5,892 5724

I 9 4,742 6,634 5988

II 10 13,673 6,850 6252

III 11 6,640 6,791 6516

IV 12 2,737 6,573 6780

I 13 3,486 6,363 7044

II 14 13,186 6,308 7308

III 15 5,448 6,932 7572

IV 16 3,485 7,887 7836

I 17 7,728 8,662 8100

II 18 16,591 8,989 8364

III 19 8,236 8628

IV 20 3,316 8892

2002

2003

2004

2005

2006

d. STEP 3 Determine the Seasonal Factors

• Formula:

• I then apply this formula to Excel by dividing columns D/F to derive to the

seasonal factor.

RESULTS:

e. STEP 4 Calculate the Average Seasonal Factors

• Savg1 = (S1+S5+S9+S13+S17)/5 = 0.76

• Savg2 = (S2+S6+S10+S14+S18)/5 = 1.90

• Savg3 = (S3+S7+S11+S15+S19)/5 = 0.95

• Savg4 = (S4+S8+S12+S16+S20)/5 = 0.41

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

I 1 3,200 3876 0.83

II 2 7,658 4140 1.85

III 3 4,420 4,472 4404 1.00

IV 4 2,384 4,657 4668 0.51

I 5 3,654 4,944 4932 0.74

II 6 8,680 5,049 5196 1.67

III 7 5,695 5,132 5460 1.04

IV 8 1,953 5,892 5724 0.34

I 9 4,742 6,634 5988 0.79

II 10 13,673 6,850 6252 2.19

III 11 6,640 6,791 6516 1.02

IV 12 2,737 6,573 6780 0.40

I 13 3,486 6,363 7044 0.49

II 14 13,186 6,308 7308 1.80

III 15 5,448 6,932 7572 0.72

IV 16 3,485 7,887 7836 0.44

I 17 7,728 8,662 8100 0.95

II 18 16,591 8,989 8364 1.98

III 19 8,236 8628 0.95

IV 20 3,316 8892 0.37

2002

2003

2004

2005

2006

f. STEP 5 Forecasting

• Finally, I forecast by multiplying the regressed factor and the seasonal factor

g. STEP 6 Error Analysis

• Error = Forecast – Demand

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft

I 1 3,200 3876 0.83 0.76 2951

II 2 7,658 4140 1.85 1.90 7862

III 3 4,420 4,472 4404 1.00 0.95 4175

IV 4 2,384 4,657 4668 0.51 0.41 1936

I 5 3,654 4,944 4932 0.74 3756

II 6 8,680 5,049 5196 1.67 9867

III 7 5,695 5,132 5460 1.04 5176

IV 8 1,953 5,892 5724 0.34 2373

I 9 4,742 6,634 5988 0.79 4560

II 10 13,673 6,850 6252 2.19 11873

III 11 6,640 6,791 6516 1.02 6177

IV 12 2,737 6,573 6780 0.40 2811

I 13 3,486 6,363 7044 0.49 5364

II 14 13,186 6,308 7308 1.80 13878

III 15 5,448 6,932 7572 0.72 7178

IV 16 3,485 7,887 7836 0.44 3249

I 17 7,728 8,662 8100 0.95 6168

II 18 16,591 8,989 8364 1.98 15884

III 19 8,236 8628 0.95 8179

IV 20 3,316 8892 0.37 3687

I 21 9156 6972

II 22 9420 17889

III 23 9684 9180

IV 24 9948 4125

2002

2003

2004

2005

2006

2007

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft Error

I 1 3,200 3876 0.83 0.76 2951 -249

II 2 7,658 4140 1.85 1.90 7862 204

III 3 4,420 4,472 4404 1.00 0.95 4175 -245

IV 4 2,384 4,657 4668 0.51 0.41 1936 -448

I 5 3,654 4,944 4932 0.74 3756 102

II 6 8,680 5,049 5196 1.67 9867 1187

III 7 5,695 5,132 5460 1.04 5176 -519

IV 8 1,953 5,892 5724 0.34 2373 420

I 9 4,742 6,634 5988 0.79 4560 -182

II 10 13,673 6,850 6252 2.19 11873 -1800

III 11 6,640 6,791 6516 1.02 6177 -463

IV 12 2,737 6,573 6780 0.40 2811 74

I 13 3,486 6,363 7044 0.49 5364 1878

II 14 13,186 6,308 7308 1.80 13878 692

III 15 5,448 6,932 7572 0.72 7178 1730

IV 16 3,485 7,887 7836 0.44 3249 -236

I 17 7,728 8,662 8100 0.95 6168 -1560

II 18 16,591 8,989 8364 1.98 15884 -707

III 19 8,236 8628 0.95 8179 -57

IV 20 3,316 8892 0.37 3687 371

2002

2003

2004

2005

2006

• Absolute deviation

• Mean squared error (MSE)

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft Error Abs. Error

I 1 3,200 3876 0.83 0.76 2951 -249 249

II 2 7,658 4140 1.85 1.90 7862 204 204

III 3 4,420 4,472 4404 1.00 0.95 4175 -245 245

IV 4 2,384 4,657 4668 0.51 0.41 1936 -448 448

I 5 3,654 4,944 4932 0.74 3756 102 102

II 6 8,680 5,049 5196 1.67 9867 1187 1187

III 7 5,695 5,132 5460 1.04 5176 -519 519

IV 8 1,953 5,892 5724 0.34 2373 420 420

I 9 4,742 6,634 5988 0.79 4560 -182 182

II 10 13,673 6,850 6252 2.19 11873 -1800 1800

III 11 6,640 6,791 6516 1.02 6177 -463 463

IV 12 2,737 6,573 6780 0.40 2811 74 74

I 13 3,486 6,363 7044 0.49 5364 1878 1878

II 14 13,186 6,308 7308 1.80 13878 692 692

III 15 5,448 6,932 7572 0.72 7178 1730 1730

IV 16 3,485 7,887 7836 0.44 3249 -236 236

I 17 7,728 8,662 8100 0.95 6168 -1560 1560

II 18 16,591 8,989 8364 1.98 15884 -707 707

III 19 8,236 8628 0.95 8179 -57 57

IV 20 3,316 8892 0.37 3687 371 371

2002

2003

2004

2005

2006

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft Error Abs. Error MSE

I 1 3,200 3876 0.83 0.76 2951 -249 249 61,773

II 2 7,658 4140 1.85 1.90 7862 204 204 51,700

III 3 4,420 4,472 4404 1.00 0.95 4175 -245 245 54,510

IV 4 2,384 4,657 4668 0.51 0.41 1936 -448 448 91,150

I 5 3,654 4,944 4932 0.74 3756 102 102 74,984

II 6 8,680 5,049 5196 1.67 9867 1187 1187 297,477

III 7 5,695 5,132 5460 1.04 5176 -519 519 293,487

IV 8 1,953 5,892 5724 0.34 2373 420 420 278,900

I 9 4,742 6,634 5988 0.79 4560 -182 182 251,604

II 10 13,673 6,850 6252 2.19 11873 -1800 1800 550,517

III 11 6,640 6,791 6516 1.02 6177 -463 463 519,970

IV 12 2,737 6,573 6780 0.40 2811 74 74 477,100

I 13 3,486 6,363 7044 0.49 5364 1878 1878 711,639

II 14 13,186 6,308 7308 1.80 13878 692 692 695,030

III 15 5,448 6,932 7572 0.72 7178 1730 1730 848,198

IV 16 3,485 7,887 7836 0.44 3249 -236 236 798,660

I 17 7,728 8,662 8100 0.95 6168 -1560 1560 894,850

II 18 16,591 8,989 8364 1.98 15884 -707 707 872,940

III 19 8,236 8628 0.95 8179 -57 57 827,167

IV 20 3,316 8892 0.37 3687 371 371 792,694

2002

2003

2004

2005

2006

• Mean absolute deviation (MAD)

• Mean absolute percentage error (MAPE)

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft Error Abs. Error MSE MAD

I 1 3,200 3876 0.83 0.76 2951 -249 249 61,773 249

II 2 7,658 4140 1.85 1.90 7862 204 204 51,700 226

III 3 4,420 4,472 4404 1.00 0.95 4175 -245 245 54,510 233

IV 4 2,384 4,657 4668 0.51 0.41 1936 -448 448 91,150 287

I 5 3,654 4,944 4932 0.74 3756 102 102 74,984 250

II 6 8,680 5,049 5196 1.67 9867 1187 1187 297,477 406

III 7 5,695 5,132 5460 1.04 5176 -519 519 293,487 422

IV 8 1,953 5,892 5724 0.34 2373 420 420 278,900 422

I 9 4,742 6,634 5988 0.79 4560 -182 182 251,604 395

II 10 13,673 6,850 6252 2.19 11873 -1800 1800 550,517 536

III 11 6,640 6,791 6516 1.02 6177 -463 463 519,970 529

IV 12 2,737 6,573 6780 0.40 2811 74 74 477,100 491

I 13 3,486 6,363 7044 0.49 5364 1878 1878 711,639 598

II 14 13,186 6,308 7308 1.80 13878 692 692 695,030 605

III 15 5,448 6,932 7572 0.72 7178 1730 1730 848,198 680

IV 16 3,485 7,887 7836 0.44 3249 -236 236 798,660 652

I 17 7,728 8,662 8100 0.95 6168 -1560 1560 894,850 705

II 18 16,591 8,989 8364 1.98 15884 -707 707 872,940 705

III 19 8,236 8628 0.95 8179 -57 57 827,167 671

IV 20 3,316 8892 0.37 3687 371 371 792,694 656

2002

2003

2004

2005

2006

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft Error Abs. Error MSE MAD % Error MAPE

I 1 3,200 3876 0.83 0.76 2951 -249 249 61,773 249 7.77 7.77

II 2 7,658 4140 1.85 1.90 7862 204 204 51,700 226 2.66 5.22

III 3 4,420 4,472 4404 1.00 0.95 4175 -245 245 54,510 233 5.55 5.33

IV 4 2,384 4,657 4668 0.51 0.41 1936 -448 448 91,150 287 18.81 8.70

I 5 3,654 4,944 4932 0.74 3756 102 102 74,984 250 2.78 7.51

II 6 8,680 5,049 5196 1.67 9867 1187 1187 297,477 406 13.68 8.54

III 7 5,695 5,132 5460 1.04 5176 -519 519 293,487 422 9.12 8.62

IV 8 1,953 5,892 5724 0.34 2373 420 420 278,900 422 21.53 10.24

I 9 4,742 6,634 5988 0.79 4560 -182 182 251,604 395 3.84 9.53

II 10 13,673 6,850 6252 2.19 11873 -1800 1800 550,517 536 13.17 9.89

III 11 6,640 6,791 6516 1.02 6177 -463 463 519,970 529 6.98 9.63

IV 12 2,737 6,573 6780 0.40 2811 74 74 477,100 491 2.72 9.05

I 13 3,486 6,363 7044 0.49 5364 1878 1878 711,639 598 53.87 12.50

II 14 13,186 6,308 7308 1.80 13878 692 692 695,030 605 5.25 11.98

III 15 5,448 6,932 7572 0.72 7178 1730 1730 848,198 680 31.75 13.30

IV 16 3,485 7,887 7836 0.44 3249 -236 236 798,660 652 6.77 12.89

I 17 7,728 8,662 8100 0.95 6168 -1560 1560 894,850 705 20.19 13.32

II 18 16,591 8,989 8364 1.98 15884 -707 707 872,940 705 4.26 12.82

III 19 8,236 8628 0.95 8179 -57 57 827,167 671 0.69 12.18

IV 20 3,316 8892 0.37 3687 371 371 792,694 656 11.19 12.13

2002

2003

2004

2005

2006

• Tracking signal (TS)

• Figure: Demand Vs Forecast using Static Method

• Static method forecasting follows demand very closely as we can see. All the tracking signal

values are inside the interval of -6 to 6 and do not go past a value of 4 meaning that this forecast

does a decent job of not under/over estimating demand too often.

Year Quarter Period

Clear Plastic

Demand ('000 lbs)

De-Seasonalized

Deamnd

Regressed

De-Seasonalized

Demand

Seasonal

Factor

Avg.

Seasonal

Factor

Forecast,

Ft Error Abs. Error MSE MAD % Error MAPE TS

I 1 3,200 3876 0.83 0.76 2951 -249 249 61,773 249 7.77 7.77 -1

II 2 7,658 4140 1.85 1.90 7862 204 204 51,700 226 2.66 5.22 -0.20

III 3 4,420 4,472 4404 1.00 0.95 4175 -245 245 54,510 233 5.55 5.33 -1.25

IV 4 2,384 4,657 4668 0.51 0.41 1936 -448 448 91,150 287 18.81 8.70 -2.58

I 5 3,654 4,944 4932 0.74 3756 102 102 74,984 250 2.78 7.51 -2.55

II 6 8,680 5,049 5196 1.67 9867 1187 1187 297,477 406 13.68 8.54 1.36

III 7 5,695 5,132 5460 1.04 5176 -519 519 293,487 422 9.12 8.62 0.08

IV 8 1,953 5,892 5724 0.34 2373 420 420 278,900 422 21.53 10.24 1.07

I 9 4,742 6,634 5988 0.79 4560 -182 182 251,604 395 3.84 9.53 0.68

II 10 13,673 6,850 6252 2.19 11873 -1800 1800 550,517 536 13.17 9.89 -2.86

III 11 6,640 6,791 6516 1.02 6177 -463 463 519,970 529 6.98 9.63 -3.77

IV 12 2,737 6,573 6780 0.40 2811 74 74 477,100 491 2.72 9.05 -3.91

I 13 3,486 6,363 7044 0.49 5364 1878 1878 711,639 598 53.87 12.50 -0.07

II 14 13,186 6,308 7308 1.80 13878 692 692 695,030 605 5.25 11.98 1.08

III 15 5,448 6,932 7572 0.72 7178 1730 1730 848,198 680 31.75 13.30 3.50

IV 16 3,485 7,887 7836 0.44 3249 -236 236 798,660 652 6.77 12.89 3.29

I 17 7,728 8,662 8100 0.95 6168 -1560 1560 894,850 705 20.19 13.32 0.83

II 18 16,591 8,989 8364 1.98 15884 -707 707 872,940 705 4.26 12.82 -0.17

III 19 8,236 8628 0.95 8179 -57 57 827,167 671 0.69 12.18 -0.27

IV 20 3,316 8892 0.37 3687 371 371 792,694 656 11.19 12.13 0.29

2002

2003

2004

2005

2006

SPC 4-point Moving Average

a. We are using a four-period moving average for this. We start by averaging every four periods to find the level. We then forecast demand by setting the next period equal to the

previous period’s level.

b. The error analysis is conducted the same way as before.

Year Quarter Period

Clear Plastic

Demand ('000 lbs) Level Lt Forecast Ft

I 1 3,200

II 2 7,658

III 3 4,420

IV 4 2,384 4,416

I 5 3,654 4,529 4416

II 6 8,680 4,785 4529

III 7 5,695 5,103 4785

IV 8 1,953 4,996 5103

I 9 4,742 5,268 4996

II 10 13,673 6,516 5268

III 11 6,640 6,752 6516

IV 12 2,737 6,948 6752

I 13 3,486 6,634 6948

II 14 13,186 6,512 6634

III 15 5,448 6,214 6512

IV 16 3,485 6,401 6214

I 17 7,728 7,462 6401

II 18 16,591 8,313 7462

III 19 8,236 9,010 8313

IV 20 3,316 8,968 9010

2002

2003

2004

2005

2006

Figure: SPC Forecasts Using Four-Period Moving Average

Year Quarter Period

Clear Plastic

Demand ('000 lbs) Level Lt Forecast Ft Error Et Abs. Error At Sq. Error MSEt MADt %Error MAPEt TSt

I 1 3,200

II 2 7,658

III 3 4,420

IV 4 2,384 4,416

I 5 3,654 4,529 4416 762 762 579,882 762 21 21 1.00

II 6 8,680 4,785 4529 -4151 4151 8,905,342 2456 48 34 -1.38

III 7 5,695 5,103 4785 -911 911 6,213,231 1941 16 28 -2.22

IV 8 1,953 4,996 5103 3150 3150 7,140,942 2243 161 61 -0.51

I 9 4,742 5,268 4996 254 254 5,725,606 1845 5 50 -0.49

II 10 13,673 6,516 5268 -8406 8406 16,546,744 2939 61 52 -3.17

III 11 6,640 6,752 6516 -124 124 14,185,128 2537 2 45 -3.72

IV 12 2,737 6,948 6752 4015 4015 14,427,016 2721 147 58 -1.99

I 13 3,486 6,634 6948 3462 3462 14,155,730 2804 99 62 -0.70

II 14 13,186 6,512 6634 -6552 6552 17,033,027 3179 50 61 -2.67

III 15 5,448 6,214 6512 1064 1064 15,587,536 2986 20 57 -2.49

IV 16 3,485 6,401 6214 2729 2729 14,909,309 2965 78 59 -1.59

I 17 7,728 7,462 6401 -1327 1327 13,897,844 2839 17 56 -2.13

II 18 16,591 8,313 7462 -9129 9129 18,858,227 3288 55 56 -4.61

III 19 8,236 9,010 8313 77 77 17,601,407 3074 1 52 -4.91

IV 20 3,316 8,968 9010 5694 5694 18,527,671 3238 172 60 -2.90

2002

2003

2004

2005

2006

Figure: SPC Forecasts Using Four-Period Moving Average

c. We forecast 2007’s demand by F21 = F22 = F23 = F24 = L20 = 8,968

d. Figure: Demand Vs. Forecast using Moving Average Method

• Moving average method does not do a great job of forecasting as you can see it doesn’t follow

the demand points. The high frequency of negative Tracking Signal values indicates it under

estimates the demand too often however it stays inside the interval of -6 to 6. This method does

not allow SPC to match the peak in clear plastic demand during summer.

Year Quarter Period

Clear Plastic

Demand ('000 lbs) Level Lt Forecast Ft Error Et Abs. Error At Sq. Error MSEt MADt %Error MAPEt TSt

I 1 3,200

II 2 7,658

III 3 4,420

IV 4 2,384 4,416

I 5 3,654 4,529 4416 762 762 579,882 762 21 21 1.00

II 6 8,680 4,785 4529 -4151 4151 8,905,342 2456 48 34 -1.38

III 7 5,695 5,103 4785 -911 911 6,213,231 1941 16 28 -2.22

IV 8 1,953 4,996 5103 3150 3150 7,140,942 2243 161 61 -0.51

I 9 4,742 5,268 4996 254 254 5,725,606 1845 5 50 -0.49

II 10 13,673 6,516 5268 -8406 8406 16,546,744 2939 61 52 -3.17

III 11 6,640 6,752 6516 -124 124 14,185,128 2537 2 45 -3.72

IV 12 2,737 6,948 6752 4015 4015 14,427,016 2721 147 58 -1.99

I 13 3,486 6,634 6948 3462 3462 14,155,730 2804 99 62 -0.70

II 14 13,186 6,512 6634 -6552 6552 17,033,027 3179 50 61 -2.67

III 15 5,448 6,214 6512 1064 1064 15,587,536 2986 20 57 -2.49

IV 16 3,485 6,401 6214 2729 2729 14,909,309 2965 78 59 -1.59

I 17 7,728 7,462 6401 -1327 1327 13,897,844 2839 17 56 -2.13

II 18 16,591 8,313 7462 -9129 9129 18,858,227 3288 55 56 -4.61

III 19 8,236 9,010 8313 77 77 17,601,407 3074 1 52 -4.91

IV 20 3,316 8,968 9010 5694 5694 18,527,671 3238 172 60 -2.90

I 21 8968

II 22 8968

III 23 8968

IV 24 89682007

2002

2003

2004

2005

2006

Figure: SPC Forecasts Using Four-Period Moving Average

SPC Simple Exponential Smoothing

• We will now perform adaptive forecasting using the method of simple exponential smoothing.

We will use a smoothing constant,  = 0.06, to smooth the forecast of the level, L. o Step 1: Initialize level

▪ L0 = average of all demand points, Di o Step 2: Initial Forecast

▪ F1 = L0; F2 =L0 o Step 3: Compute the forecast error

▪ E1 = F1 – D1 = (L0 – D1) o Step 4: Modification, adapt the level based on forecast error If E1 > 0, F1 > D1 and thus

we are over predicting the demand. Therefore, to improve the forecast, we should

(from eq.(3)) reduce the level.

▪ L1 = L0 - E1 o Combining equations from step 3 and step 4 we get

▪ L1 = D1 + (1-)L0 ▪ Forecast F2 = L1

o Our general equations are: ▪ Ft+1 = Lt ▪ Lt+1 = Dt+1 + (1-)Lt

o The demand forecast, ▪ Ft+i = Lt+i (i = 2,3,4…)

• Forecast for Clear Plastic using Simple Exponential Smoothing

Year Quarter Period

Clear Plastic

Demand ('000 lbs) Level Lt Forecast Ft Error Et Abs. Error At Sq. Error MSEt MADt %Error MAPEt TSt

0 0 6,346

I 1 3,200 6,157 6346 3146 3146 9,894,799 3146 98 98 1.00

II 2 7,658 6,247 6157 -1501 1501 6,074,104 2323 20 59 0.71

III 3 4,420 6,137 6247 1827 1827 5,161,963 2158 41 53 1.61

IV 4 2,384 5,912 6137 3753 3753 7,393,318 2557 157 79 2.83

I 5 3,654 5,777 5912 2258 2258 6,934,473 2497 62 76 3.80

II 6 8,680 5,951 5777 -2903 2903 7,183,654 2565 33 69 2.57

III 7 5,695 5,935 5951 256 256 6,166,767 2235 4 59 3.06

IV 8 1,953 5,697 5935 3982 3982 7,378,442 2453 204 78 4.41

I 9 4,742 5,639 5697 955 955 6,659,853 2287 20 71 5.15

II 10 13,673 6,121 5639 -8034 8034 12,447,963 2862 59 70 1.31

III 11 6,640 6,152 6121 -519 519 11,340,790 2649 8 64 1.22

IV 12 2,737 5,947 6152 3415 3415 11,367,809 2712 125 69 2.45

I 13 3,486 5,800 5947 2461 2461 10,959,432 2693 71 69 3.38

II 14 13,186 6,243 5800 -7386 7386 14,073,474 3028 56 68 0.56

III 15 5,448 6,195 6243 795 795 13,177,374 2879 15 65 0.87

IV 16 3,485 6,033 6195 2710 2710 12,812,886 2869 78 66 1.82

I 17 7,728 6,134 6033 -1695 1695 12,228,257 2800 22 63 1.26

II 18 16,591 6,762 6134 -10457 10457 17,623,411 3225 63 63 -2.15

III 19 8,236 6,850 6762 -1474 1474 16,810,250 3133 18 61 -2.68

IV 20 3,316 6,638 6850 3534 3534 16,594,275 3153 107 63 -1.55

I 21 6638

II 22 6638

III 23 6638

IV 24 66382007

Figure: SPC Forecasts Using Simple Exponential Smoothing

2002

2003

2004

2005

2006

• Figure: Demand Vs. Forecast using Simple Exponential Smoothing

• As is clear from our plots we can visually see that simple exponential smoothing is not an

accurate forecasting method for this data. This is confirmed by the large error metrics MAPE and

MAD in tables above.

• Simple Exponential Smoothing does not do a great job at forecasting demand. It under and over

estimates demand at many points throughout the forecast. As the forecast progresses, the

simple exponential smoothing error values become higher meaning that this forecast is highly

inaccurate and unreliable. Tracking Signal at period 9 is equal to 5.15 meaning that this forecast

was close to breaking past the highest value of 6 almost making this forecast biased.

Holt’s Method

• We now forecast demand using Level and Trend corrected exponential smoothing. The

assumption is that the data has level, L, and trend, T, only.

• Process:

o Step 1: Regress the given data to compute the initial values of the level, L0, and initial

trend T0.

▪ Forecast, F1 = L0 + T0

o Step 2: Adapt Use two smoothing constants, =0.06 and =0.06, to smooth respectively

level and trend.

▪ L1 = D1 + (1-)[L0 + T0]

▪ T1 = [L1 – L0] + (1-)T0

▪ Forecast, F2 = L1 + T1

o Step 3: Forecast

▪ Ft+1 = Lt + Tt

▪ Lt+1 = Dt+1 + (1-)[Lt + Tt]

▪ Tt+1 = [Lt+1 – Lt] + (1-)Tt

o In order to obtain L0 and T0, I graphed the Demand for clear plastic and obtained the

slope of the functions

o L0 = 4134 and T0 = 211

o After applying all the above equations into our spreadsheet, we arrive at the following:

o To forecast for 2007, we use the formulas:

▪ F21 = L20 + T20 = 8356 + 211 = 8,567

▪ F22 = L20 + 2*T20 = 8356 + 2*211 = 8,778

Year Quarter Period

Clear Plastic

Demand ('000 lbs) Level Lt Trend Tt Forecast Ft Error Et Abs. Error At Mean Sq. Error MSEt MADt %Error MAPEt TSt

0 0 4,134 211

I 1 3,200 4,276 207 4,345 1,145 1145 1,311,025 1145 35.78 35.78 1

II 2 7,658 4,674 218 4,483 -3,175 3175 5,695,260 2160 41.46 38.62 -0.94

III 3 4,420 4,864 217 4,892 472 472 3,871,093 1597 10.68 29.31 -0.98

IV 4 2,384 4,918 207 5,080 2,696 2696 4,720,780 1872 113.10 50.25 0.61

I 5 3,654 5,037 202 5,125 1,471 1471 4,209,622 1792 40.27 48.26 1.46

II 6 8,680 5,445 214 5,239 -3,441 3441 5,481,763 2067 39.65 46.82 -0.40

III 7 5,695 5,661 214 5,659 -36 36 4,698,837 1777 0.63 40.22 -0.49

IV 8 1,953 5,640 200 5,875 3,922 3922 6,034,688 2045 200.84 60.30 1.49

I 9 4,742 5,774 196 5,840 1,098 1098 5,498,149 1940 23.16 56.17 2.14

II 10 13,673 6,432 224 5,970 -7,703 7703 10,881,540 2516 56.34 56.19 -1.41

III 11 6,640 6,655 224 6,656 16 16 9,892,333 2289 0.24 51.10 -1.54

IV 12 2,737 6,630 209 6,879 4,142 4142 10,497,623 2443 151.33 59.46 0.25

I 13 3,486 6,638 197 6,839 3,353 3353 10,555,059 2513 96.19 62.28 1.58

II 14 13,186 7,216 220 6,835 -6,351 6351 12,682,401 2787 48.17 61.27 -0.86

III 15 5,448 7,316 212 7,435 1,987 1987 12,100,242 2734 36.48 59.62 -0.15

IV 16 3,485 7,286 198 7,529 4,044 4044 12,365,928 2816 116.03 63.15 1.29

I 17 7,728 7,499 199 7,484 -244 244 11,642,025 2665 3.16 59.62 1.28

II 18 16,591 8,231 231 7,697 -8,894 8894 15,389,528 3011 53.61 59.28 -1.83

III 19 8,236 8,448 230 8,462 226 226 14,582,236 2864 2.74 56.31 -1.84

IV 20 3,316 8,356 211 8,678 5,362 5362 15,290,771 2989 161.71 61.58 0.03

I 21 8,567

II 22 8,778

III 23 8,988

IV 24 9,199

Figure: SPC Forecasts Using Holt's Method

2002

2003

2004

2005

2006

2007

▪ F23 = L20 + 3*T20 = 8356 + 3*211 = 8,988

▪ F24 = L20 + 4*T20 = 8356 + 4*211 =9,199

o Figure: Demand vs. Forecast for Clear Plastic using Holt’s Method

o As we can see from Figure above, the forecast does not very accurately represent actual

demand. The high % error, MAPE, and MAD in Tables above confirm that this is not an

accurate forecasting method for this data set.

Winter’s Method

• Winter’s Method starts off similarly to the static method. We de-seasonalize demand by running

it through a regression analysis, and then find the seasonal factors for it. Since the data remains

the same as the static method, the values are equal to: o L = 3,612 and T = 264 o Savg1 = (S1+S5+S9+S13+S17)/5 = 0.76 o Savg2 = (S2+S6+S10+S14+S18)/5 = 1.90 o Savg3 = (S3+S7+S11+S15+S19)/5 = 0.95 o Savg4 = (S4+S8+S12+S16+S20)/5 = 0.41

• Initial Forecast o F1 = (L0 + T0)(S1) = 2,946

• Adaptation o Let =0.06, =0.06, =0.06

▪ Lt+1 = (Dt+1/St+1) + (1-)(Lt + Tt)

▪ Tt+1 = (Lt+1 – Lt) + (1 - )Tt

▪ St+p+1 = (Dt+1/Lt+1) + (1 - )St+1 o Applying above 4 equations to our spreadsheet yields:

• To forecast demand for 2007 (the next four quarters/periods)

o S21 = 0.76, S22 = 1.90, S23 = 0.95, S24 = 0.41

o F21 = (L20 + T20)*S21 = (8,857 + 262)*0.76 = 6,930

o F22 = (L20 + 2*T20)*S22 = (8,857 + 2*262)*0.76 = 17,823

o F23 = (L20 + 3*T20)*S23 = (8,857 + 3*262)*0.76 = 9,160

o F24 = (L20 + 4*T20)*S24 = (8,857 + 4*262)*0.76 = 4,061

o Figure: Demand Vs Forecast using Winter’s Method

o The Winter’s method forecast similarly resembles the static method forecast. It closely

follows the demand points which means it did a great job at forecasting. Having a trend

factor is important in determining a more accurate forecast for future demand sales.

Year Quarter Period

Clear Plastic

Demand ('000 lbs) Level Lt Trend Tt Seasonal Factors, Si Forecast Ft Error Et Abs. Error At Mean Sq. Error MSEt MADt %Error MAPEt TSt

0 0 3,612 264

I 1 3,200 3,896 265 0.76 2,946 -254 254 64,638 254 7.94 7.94 -1

II 2 7,658 4,153 265 1.90 7,906 248 248 63,176 251 3.24 5.59 -0.02

III 3 4,420 4,432 266 0.95 4,197 -223 223 58,656 242 5.04 5.41 -0.95

IV 4 2,384 4,765 270 0.41 1,926 -458 458 96,409 296 19.21 8.86 -2.32

I 5 3,654 5,019 269 0.76 3,845 191 191 84,400 275 5.22 8.13 -1.80

II 6 8,680 5,245 266 1.90 10,030 1,350 1350 373,912 454 15.55 9.37 1.88

III 7 5,695 5,539 268 0.95 5,252 -443 443 348,579 452 7.79 9.14 0.91

IV 8 1,953 5,741 264 0.42 2,412 459 459 331,396 453 23.53 10.94 1.92

I 9 4,742 6,018 265 0.76 4,573 -169 169 297,753 422 3.57 10.12 1.66

II 10 13,673 6,342 268 1.88 11,825 -1,848 1848 609,550 564 13.52 10.46 -2.03

III 11 6,640 6,629 269 0.96 6,328 -312 312 562,992 541 4.70 9.94 -2.70

IV 12 2,737 6,884 268 0.41 2,835 98 98 516,872 504 3.57 9.41 -2.70

I 13 3,486 6,998 259 0.76 5,459 1,973 1973 776,421 617 56.59 13.03 0.99

II 14 13,186 7,238 258 1.90 13,778 592 592 745,981 616 4.49 12.42 1.95

III 15 5,448 7,387 251 0.96 7,197 1,749 1749 900,076 691 32.10 13.74 4.27

IV 16 3,485 7,690 255 0.41 3,133 -352 352 851,581 670 10.11 13.51 3.88

I 17 7,728 8,089 263 0.75 5,936 -1,792 1792 990,281 736 23.18 14.08 1.10

II 18 16,591 8,376 265 1.89 15,818 -773 773 968,477 738 4.66 13.56 0.05

III 19 8,236 8,645 265 0.95 8,180 -56 56 917,669 702 0.68 12.88 -0.03

IV 20 3,316 8,857 262 0.41 3,677 361 361 878,296 685 10.88 12.78 0.49

I 21 0.76 6,930

II 22 1.90 17,823

III 23 0.95 9,160

IV 24 0.41 4,0612007

Figure: SPC Forecasts Using Winter's Method

2002

2003

2004

2005

2006

o Which forecasting method should Julie Williams use for Clear Plastic?

▪ After analyzing all of the above figures and tables, we conclude that Julie should

use Winter’s forecasting method to forecast clear plastic. We arrive at this

conclusion from observing that this method yields the lowest values of MAD and

MAPE for clear plastics. This is also verified visually in Figure: “Winter’s Method:

Demand Vs. Forecast”. Intuitively this result also makes sense seeing as Winter’s

method corrects for level, trend, and seasonality. Interestingly, the basic

method of static forecasting is almost as good a forecast for this data set. Upon

analyzing the data, we can see that this too makes sense since the greatest

factor in the data is seasonality and that there is not a very significant trend.

o Was my hypothesis correct? ▪ Yes, my hypothesis was correct because Winter’s method was the best for

Julie’s data because it took into account both trend and seasonality that the

demand had shown. With Winter’s Method Julie should be able to improve her

supply chain and match the demand with supply.

o 2007 Demand Forecast for Clear Plastic (‘000 lb.):

o The annual Demand for the year 2007 is 37,974.

4. Check your work a. Is the work correct in every detail?