Chapter 1: Introduction and Productivity

Strategic Decision: longer term decision by senior management

• product/service strategy
• competitive priorities
• location

Tactical Decision: medium term decisions by middle and senior managers

• product/process design
• job design and workforce management
• capacity management

Operating Operations: short term decisions by middle and lower management

• forecasting
• material/inventory management
• production control

The core functions of organizations include three parts:

1) Marketing: generates demand, sell, assess customer wants.

2) Production/Operations: creates the product, adds value.
3) Finance/Accounting: pays bills, collects money, budgets, and provides funds for operations.

• the challenge of the operation manager is “matching the supply with demand”.

Operations management can help 1) improve efficiency 2) make trade-offs 3) invest in new technologies.

Outputs
Multi-factor Productivity =
Inputs

• Keep note of what the output is given in (daily, weekly, monthly)

• If given output for the month, make sure all of your inputs match the output.
• For example, Output = 10000 units a month, then your days worked and hours must be per
month, not daily/week.

—> just multiply # days worked per week by 4 to get the total worked for the month

New Production−Old Production

% Change In Productivity = x 100
Old Production

Example: A bank employs 3 loan officers, each working 8 hours per day. Each officer processes an
average of 5 loans per day. The bank’s payroll cost for the officers is $800 per day, and there is a daily
overhead expense of $500. The bank has just purchased new computer software that should enable each
officer to process 8 loans per day, although the overhead expense will increase to $550.

Calculate change in labour productivity. Calculate change in multi-factor productivity.

Solution:

** make sure you pay attention to what the question is asking (ie. productivity in loans per hour, or
loans per dollar **

MODULE E: LEARNING CURVES

Calculating Learning Curves

Note: probably will never have to use this, but:
Arithmetic Analysis = TN = T1 x Ln Question: “If the first unit of a particular product
took 20 labour hours, how long will (a) the eighth
unit and (b) the fourth unit take if you are on a
75% learning curve rate?”

Answer: 8th unit doubles 3 times (1 to 2, 2 to 4,

and 4 to 8).
 - TN = total time for the n-th unit (hours)
- T1 = time of the 1st unit (hours)
- L = learning curve rate
- N = # times T1 is doubled

Logarithmic Analysis = TN = T1 (Nb)

Note: Will probably just have a multiple choice question
- TN = total time for the n-th unit about this.
- T1 = hours to produce 1st unit
log(L) Question: If Professor TQ takes 20 minutes to grade the first
- B= ; L = learning curve exam and follows an 75% learning curve, how long will it
log(2)
take her to grade: 1. the 25th exam?
rate in decimal form
Answer: Always find your b-value first

B = log (.75) / log (2) = -.415

Then, convert time for the first unit into hours (if given in
20 Min
minutes), and plug everything into the equation = =
60 Min
.33

Learning-Curve Coefficients = TN = T1 (C) Note: definitely going to be used to calculate the

most, but it will tell you which method to use in
- TN = total time for the n-th unit
the question.
- T1 = hours to produce the 1st unit
- C = coefficient (found on table E.3) - Will probably be a short answer question!

Scenario 1

Question: “How long did it take to make the 4th unit”:

Answer: Go to your Table E.3, look up n = 4 and look under “Unit time Coefficient”. Multiply
the time it takes for the 1st unit (in hours, which will be given), by the coefficient you find. This
will give you the time it takes to make the 4th unit.
Scenario 2

Question: “How long does it take to make the first 4 units”:

Answer: Go to your Table E.3, look up n = 4 and look under “Total Time Coefficient”.
Multiply the time it takes for the 1st unit (in hours, which will be given), by the coefficient you
find. This will give you the time it takes to make the first 4 units.
Scenario 3

Question: “How long does it take to make between 10 to 20 units?”

Answer: Go to your Table E.3, look up n = 10, and n = 20 and look under “Unit Time
Coefficient”. Multiply the time it takes for the 1st unit (in hours, which will be given), by the
coefficient you find for these values.
Then, do your normal n-th unit calculations for when n = 10 and n = 20.
Once you have the time it takes to make the 10th and the 20th unit, you simply subtract T20 – T10,
and that will give you the time it takes to make between 10 and 20 units.

Scenario 4

Question: “How much does it cost to make the first 20 units? The company pays its employees
$20.00 an hour”

Answer: Go to your Table E.3, look up n = 20 and look under “Total Time Coefficient”.
Multiply the time it takes for the 1st unit (in hours, which will be given), by the coefficient you
find. This will give you the total time it takes (in hours) to make the first 20 units.

Then to answer the question, simply multiply the total hours you calculated by the wage per hour
that is given in the question.

Progress Rate = 1 – Learning Curve Rate (decimal)

CHAPTER 3: PROJECT MANAGEMENT

Critical path = the longest connected path through the network with the most weeks.

= A-B-D-E-G-I-J-K
= 4+6+6+14+2+3+4+2

Therefore, the expected project length is 41 weeks.

Slack = the maximum length of time that an activity can be delayed without delaying the entire
project. Thus, activities on the critical path have zero slack.
- Slack = Latest Start (LS) – Earliest Start (ES); or Latest Finish (LF) – Earliest Finish (EF)

Expected Time, t, Variance, and Standard Deviation

(a +4 m+b)
t= a = optimistic time; b = pessimistic time; m = most likely time
6

b−a 2
Variance (σ )2 = ( ¿
6
Standard Deviation = √(σ )2

Z score

TS = due date; always given in the question (ie. Finish in less than 20 weeks)
Te = expected completion time  calculated my checking the critical path
σ = standard deviation if not given in the question, just square root the total
variance of the critical path. (ie. Critical path is A-C-E and variance for a is 8, c is 2,
 Ts−Te
Z=
σ

Scenario 1 (after they ask you to calculate variance, expected completion time, etc.)

Question: What is the probability the project will finish in less than 20 weeks?

Answer: You get your Ts, Te, and your standard deviation, and plug them into the equation.
The z-score that you get will give you the percentage you need for the probability that they will
finish in less than 20 weeks (ie. A 1.85 z score = .96784 – 97%).

Scenario 2

Question: What is the probability the project will finish in more than 20 weeks?

Answer: Calculate your z score as usual.

Look up the % corresponding to your z-score (ie. 1.85 z score = .96784 – 97%). Then since it’s
asking for more than 20 weeks, you take your .96784 and subtract it from 1.
So your answer for finishing in more than 20 weeks will be (1 - .96784 = .03216 – 3%).

Scenario 3

Question: What is the probability the project is finished between weeks 17 and 20?

Answer: Calculate your z scores corresponding to when Ts = 17 and Ts = 20.

Find the % on the table, then subtract the percentage of the lower number of week (in this
case 17), from the percentage you get from the higher number of week (20).
So say the z score for 20 weeks is 1.25 (which has % of .89435), and the z score for week 17 is
1.00 (which has % of .84134), the probability that the project is finished between weeks 17 and
20 is: .89435 - .84134 = .05301, or 5%.

**Note: if they give you a question that gives you a mean, it is used as your Te **

CHAPTER 4: FORECASTING

METHOD OF FORMULA VARIABLES NOTES

FORECASTING
 Naïve F2010 = A2009 F2010 = forecast for 2010 - The forecast being calculated
A2009 = actual for 2009 is always going to be the
actual value of the previous
year.
- Ie. F2010 will equal the A2009.

Simple Moving F2010 = forecast for 2010 - Pay attention to how many
Average periods the question wants
A 2009+ A 2008 A2009 = actual for 2009 - For the actual values, you
F2010 = always use the previous
n n = however many periods it actual values (ie. If
asks for (two period n = 2, three calculating F2010, use actual
period n = 3) values for 2009 and 2008).

Weighted A1 = actual of oldest year - A1/W1 correspond to the

Moving Average oldest year. So if you’re
W1= weight of oldest year calculating F2010, A1 would be
the actual value for 2008.
F2010 = A2 = actual of most recent year - A2/W2 correspond to the
( A 1 x W 1 )+( A 2 x W 2) most recent year. So if
W2 = weight of most recent year you’re calculating F2010, your
W 1+W 2 A2 would be the actual value
for 2009.

Exponential - ∝ can only be between 0

Smoothing and 1
- The F and A values that are
used in the calculation are
the previous year’s values.
F2010 = F2009 + ∝(A2009 – F2009) ∝ = alpha value given in - So if you’re calculating the
question forecast for 2010, your
formula will be:
F2010 = F2009 + ∝ (A2009 – F2009)

**Note: when calculating MAD and MSE, it may ask you which ∝ value is better. It will always
be the one with the lowest MAD value**

CHAPTER 6: QUALITY MANAGEMENT

Three-Sigma/Six-Sigma: acts as the standard deviation for some of your calculations.

- Sigmas are used to reduce defects, lower costs, and improve customer satisfaction.
Chart, formula, variable, when to use, notes

Type of Formula Variables When To Use Calculating Variables

Control Chart

x́ = mean of the x́ = Look at picture

sample means below
X-bar z = sigma that is
given (ie. 3,6, etc.)
(Population UCLx = x́ + zσ x σ
σ x = population σx = ; where σ will
(process) σ is √n
Known LCLx = = x́ - zσ x standard
be given in the
deviation.
question, and n is the #
of samples.

**you will have to

calculate every time
for x chart**

X-bar x́ = mean of
(Population sample means
(process) σ is UCLx = x́ + A2 R A2 = value from
Unknown) the table
LCLx = x́ - A2 R R = average
range of the
samples
D4 = value from R = look at picture
table. below
UCLR = D4 R D3 = value from
R-Chart table.
LCLR = D3 R R = average of
sample ranges.

p = average - It will talk p=

fraction of about Total ¿ Defectives ¿ Obs
defects in the proportions of Total ¿
UCLp = p + zσ^ p population defective items
P-Chart z = sigma from a process. p(1− p)
σ^ p = √
LCLp = p - zσ^ p σ p = standard n
deviation of
sample z = given in question;
proportion. says “construct a 3
sigma control chart”, so
z = 3.
c = the mean - Use when the ¿ of defects
c=
number of goal is to ¿ of samples
 defects per control the # of
UCLc = c + z√ c unit. defects per z = given in question;
C-Chart z = sigma unit says “construct a 3
LCLc = c - z√ c - For example, sigma control chart”, so
√ c = standard one roll has 7 z = 3.
deviation of the defects in it.
defects

Note1: p-chart and c-chart are very similar. You will need to memorize the fact that you use p-
chart when it talks about proportions. If not, then use c-chart formulas. That’s how I could
tell the difference between the two.

Note2: if you ever get a negative LCL, you must limit to 0. So if you calculate an LCL of -1.34,
your answer for the LCL = 0.

Note3: anything that falls in the UCL and LCL range are considered “in control”.

x́ 

R  To get R , just divide the ranges (R) you get by the # of samples (in this case 5).