12/22/2021

INVENTORY MANAGEMENT
Vinay Vasudev, Ph.D.

MBAD 6141M – 084 OPERATIONS MANAGEMENT

What is Inventory?

1
 12/22/2021

Inventory

 A stock or store of goods

 Independent demand items
o Items that are ready to be sold or used
o Inventories are a vital part of business
 Necessary for operations
 Contribute to customer satisfaction

 A “typical” firm has roughly 30% of its

current assets and as much as 90% of its
working capital invested in inventory

Winter - 2022 MBAD 6141M - 084 3

Inventory Functions

 Inventories serve a number of functions such as:

1. To meet anticipated customer demand
2. To smooth production requirements
3. To decouple operations
4. To protect against stockouts
5. To take advantage of order cycles
6. To hedge against price increases
7. To permit operations
8. To take advantage of quantity discounts

Winter - 2022 MBAD 6141M - 084 4

2
 12/22/2021

Types of Inventory

 Raw materials and purchased parts

 Work-in-process (WIP)
 Finished goods inventories or merchandise
 Tools and supplies
 Maintenance and repairs (MRO) inventory
 Goods-in-transit to warehouses or customers (pipeline
inventory)

Winter - 2022 MBAD 6141M - 084 5

Objectives of Inventory Control

 Inventory management has two main concerns:

1. Level of customer service
o Having the right goods available in the right quantity in the
right place at the right time
2. Costs of ordering and carrying inventories
o The overall objective of inventory management is to achieve
satisfactory levels of customer service while keeping
inventory costs within reasonable bounds
1. Measures of performance
2. Customer satisfaction
 Number and quantity of backorders
 Customer complaints
3. Inventory turnover
Winter - 2022 MBAD 6141M - 084 6

3
 12/22/2021

Inventory Management

 Management has two basic functions concerning inventory:

1. Establish a system for tracking items in inventory
2. Make decisions about
o When to order?
o How much to order?
 Effective Inventory Management Requires:
1. A system to keep track of inventory
2. A reliable forecast of demand
3. Knowledge of lead time and lead time variability
4. Reasonable estimates of holding, ordering and shortage costs
5. A classification system for inventory items

Winter - 2022 MBAD 6141M - 084 7

Inventory Counting Systems

 Periodic System
o Physical count of items in inventory made at periodic
intervals
 Perpetual Inventory System
o System that keeps track of removals from inventory
continuously, thus monitoring current levels of each
item
 An order is placed when inventory drops to a
predetermined minimum level
 Two-bin system
 two containers of inventory
 reorder when the first is empty

Winter - 2022 MBAD 6141M - 084 8

4
 12/22/2021

Demand Forecasts and Lead Time

 Forecasts
o Inventories are necessary to satisfy customer demands, so it is important
to have a reliable estimates of the amount and timing of demand
o Point-of-sale (POS) systems
 A system that electronically records actual sales
 Such demand information is very useful for enhancing forecasting and inventory
management

 Lead time
o Time interval between ordering and receiving the order

Winter - 2022 MBAD 6141M - 084 9

Inventory Costs

 Purchase cost
o The amount paid to buy the inventory items
 Holding (carrying) costs
o Cost to carry an item in inventory for a length of time, usually a year
o Obsolescence; Insurance; Interest; Pilferage; Damage, Warehousing; etc.
 Ordering costs
o Costs of ordering and receiving inventory
o Supplies; Forms; Order Processing; Clerical Support; etc.
 Setup costs
o The costs involved in preparing equipment for a job
o Analogous to ordering costs
o Clean-up costs; Re-tooling costs; Adjustment costs; etc.
 Shortage costs
o Costs resulting when demand exceeds the supply of inventory; often unrealized
profit per unit

Winter - 2022 MBAD 6141M - 084 10

10

5
 12/22/2021

Inventory Carrying Costs

(Approximate Ranges)
Cost as a
Category % of Inventory Value
Housing costs (building rent, depreciation, 6%
operating cost, taxes, insurance) (3 - 10%)

Material handling costs (equipment, lease 3%

or depreciation, power, operating cost) (1 - 3.5%)
Labor cost from extra handling 3%
(3 - 5%)
Investment costs (borrowing costs, taxes, 11%
and insurance on inventory) (6 - 24%)

Pilferage, scrap, and obsolescence

3%
(2 - 5%)
Overall carrying cost 26%

Winter - 2022 MBAD 6141M - 084 11

11

ABC Classification System

 Classifying inventory according to some measure of importance, and

allocating control efforts accordingly
o A items (very important)
 10 to 20 percent of the number of items in inventory and about 60 to 70 percent of
the annual dollar value
 Counted daily/weekly
o B items (moderately important)
 Typically 25-30% of items
 Counted monthly
o C items (least important) % Annual $ Usage Class % $ Vol % Items
100 A 80 15
 Counted semi-annually/yearly B 15 30
80 C 5 55
 50 to 60 percent of the number 60
of items in inventory but only A
40
about 10 to 15 percent of the B
20 C
annual dollar value 0
0 50 100
% of Inventory Items

Winter - 2022 MBAD 6141M - 084 12

12

6
 12/22/2021

Cycle Counting

 Cycle counting
o A physical count of a sample of total inventory on a regular basis
o Used often with ABC classification
 Cycle counting management
o How much accuracy is needed? – APICS recommends
 A items: ± 0.2 percent
 B items: ± 1 percent
 C items: ± 5 percent
o When should cycle counting be performed?
o Who should do it?

Winter - 2022 MBAD 6141M - 084 13

13

Economic Order Quantity

Models

14

7
 12/22/2021

How Much to Order: EOQ Models

 Economic order quantity (EOQ) models identify the

optimal order quantity by minimizing the sum of
annual costs that vary with order size and frequency
1. The basic economic order quantity model
2. The economic production quantity model
3. The quantity discount model

Winter - 2022 MBAD 6141M - 084 15

15

Basic EOQ Model

 The basic EOQ model is used to find a fixed order quantity that
will minimize total annual inventory costs
 Assumptions:
1. Only one product is involved
2. Annual demand requirements are known
3. Demand is even throughout the year
4. Lead time does not vary
5. Each order is received in a single delivery
6. There are no quantity discounts

Winter - 2022 MBAD 6141M - 084 16

16

8
 12/22/2021

The Inventory Cycle

Profile of Inventory Level Over Time

Q Usage
Quantity rate
on hand

Reorder
point

Time
Receive Place Receive Place Receive
order order order order order

Lead time

Winter - 2022 MBAD 6141M - 084 17

17

Minimized Total Annual Cost

Total Cost  Annual Holding Cost  Annual Ordering Cost
2 DS
Q D
 H  S Qo 
2 Q
H
where
Q  Order quantity in units
H  Holding (carrying) cost per unit, usually per year
2 ( annual demand)(or der cost)

D  Demand, usually in units per year
annual per unit holding cost
S  Ordering cost per order

The Total-Cost Curve is U-Shaped

Q D D
TC  H  S
Expected number N 
Q
Annual Cost

2 Q of orders o

Holding Costs
Expected time WorkingDays / Year
T 
between orders N
Ordering Costs
Order Quantity (Q)
QO (optimal order quantity)

Winter - 2022 MBAD 6141M - 084 18

18

9
 12/22/2021

Example: Basic EOQ

A toy manufacturer uses approximately 32,000 silicon chips annually.

The chips are used at a steady rate during the 240 days a year that the
plant operates. Annual holding cost is $3 per chip, and ordering cost
is $120.
a. The optimal order quantity
D = 32,000 chips/year S = $120 per order
H = $3 per unit per year
2 DS 2 X 32000 X 120
Qo    1, 600 chips
H 3

b. The number of weekdays in an order cycle

Number of weekdays in Q
o 1600 1 1
an order cycle T    year  X 240 days  12 days
D 32000 20 20

Winter - 2022 MBAD 6141M - 084 19

19

Example: Basic EOQ

A large law firm uses an average of 40 packages of copier paper a

day. The firm operates 260 days a year. Storage and handling costs
for the paper are $30 a year per pack, and it costs $60 to order and
receive a shipment of paper. The order lead time is 3 days.
a. The optimal order quantity
D = 40 packages/day x 260 days/year = 10, 400 packages/year
S = $60 per order H = $30 per package per year LT = 3 days

2 DS 2( 10400 )60
Qopt    203 .96  204 packages .
H 30

Winter - 2022 MBAD 6141M - 084 20

20

10
 12/22/2021

Example: Basic EOQ (cont’d.)

b. The annual ordering costs

D 10,400
Annual ordering cost  S  60  $3,058.82.
Qopt 204
c. The annual holding costs
Qopt 204
Annual holding cost  H  30  $3,060.
2 2
d. Total annual costs(ordering + holding) if EOQ is ordered
D Qopt
TC  S H
Qopt 2
 $3,058.82  $3,060  $ 6,118.82.

OR TC  2 DSH  2  10,400  60  30  $6,118.82.

Winter - 2022 MBAD 6141M - 084 21

21

Example: Basic EOQ (cont’d.)

e. How many time per year an order be placed?

D 10,400
  51 / year.
Qopt 204

f. What is the reorder point?

R  dLT  40 packages/d ay  3 days  120 packages.

g. What is the expected time between orders?

WorkingDay s / Year 260
T    5. 1 days
D / Qopt 10400 / 204

Winter - 2022 MBAD 6141M - 084 22

22

11
 12/22/2021

Economic Production Quantity

23

Economic Production Quantity (EPQ)

 The batch mode is widely used in production. In certain

instances, the capacity to produce a part exceeds its usage
(demand rate)
o Assumptions
1. Only one item is involved
Q
2. Annual demand requirements are
Production Usage Production Usage Production
known Qp and usage only and usage only and usage

3. Usage rate is constant Cumulative

production
Imax
4. Usage occurs continually, but
production occurs periodically
5. The production rate is constant Amount
on hand
6. Lead time does not vary
Time
7. There are no quantity discounts

Winter - 2022 MBAD 6141M - 084 24

24

12
 12/22/2021

EPQ – Total Cost

TCmin Carrying Cost Setup Cost

𝐼 𝐷
𝐻 𝑆
2 𝑄
where
Maximum inventory 𝐼
𝑄
𝑝 𝑢
𝑝
𝑝 Production or delivery rate
𝑢 Usage rate

2 DS p
Qp 
H p u
Winter - 2022 MBAD 6141M - 084 25

25

Example: Basic EPQ

The Dine Corporation is both a producer and a user of brass

couplings. The firm operates 220 days a year and uses the
couplings at a steady rate of 50 per day. Couplings can be
produced at a rate of 200 per day. Annual storage cost is $2 per
coupling and machine setup cost is $70 per run.
a. Determine the economic run quantity
D = 50 units per day X 220 days per year = 11,000 units per year
S = $70 per order H = $2 per unit per year
p = 200 units per day u = 50 units per day
2 DS p 2 x11000 x 70 200
Qp  X  X  1013 units
H pu 2 200  50

Winter - 2022 MBAD 6141M - 084 26

26

13
 12/22/2021

Example: Basic EPQ (cont’d.)

b. Number of runs per year

D / Q p  11000 / 1013  10 .86  11 runs/year

c. Maximum units in inventory

Qp 1013
Imax  ( p  u)   ( 200  50)  759.75  760 units
p 200
c. Length of various cycles
Qp
1013 units
Length of total cycle   20.26 days
u
50 units/day
Qp 1013 units
Length of production run    5.065 days
p 200 units/day
Length of pure consumptio n period  Length of total cycle - Length of production run
 20.26 - 5.065  15.20 days

Winter - 2022 MBAD 6141M - 084 27

27

Quantity Discount Model

 Quantity discount
Adding
o Price reduction for larger PD does
not
orders offered to customers change
EOQ
to induce them to buy in
large quantities

The total-cost
Total Cost Carrying Cost Ordering Cost Purchasing Cost curve with
quantity
discounts is
𝐻 𝑆 𝑃𝐷 composed of a
portion of the
where 𝑃 Unit price total-cost curve
for each price

Winter - 2022 MBAD 6141M - 084 28

28

14
 12/22/2021

Example: Quantity Discount

A small manufacturing firm uses roughly 3400 pounds of chemical dye a year.
Currently the firm purchases 300 pounds per order and pays $3 per pound. The
supplier has just announced that orders of 1000 pounds or more will be filled at a
price of $2 per pound. The manufacturing firm incurs a cost of $100 each time it
submits an order and assigns an annual holding cost of 17 percent of the purchase
price per pound.
a. Determine the order size that will minimize the total cost.
D = 3,400 pounds per year S = $100 per order H = $0.17P
Quantity ranges are:
 1 – 999 @ unit price of $3
 1000+ @ unit price of $2

2 DS 2( 3400 )100
Q$2 / pound    1414 pounds
H 0.17 ( 2)
Because this quantity is feasible at $2 per pound, it is the optimum.

Winter - 2022 MBAD 6141M - 084 29

29

Example: Quantity Discount (cont’d.)

b. If the supplier offered the discount at 1,500 pounds instead of 1,000

pounds, what order size would minimize the total cost.
2 DS 2( 3400 )100
Q$3 / pound    1,155 pounds
H 0.17 (3)
D Q
TC  S H  PD
Q 2
3400 1155
TC 1155  100  0.17 x3  3 x3400  $10,789
1155 2

3400 1500
TC 1500  100  0.17 x 2  2 x3400  $7,282
1500 2

Lower total cost  order size of 1500 pound is the optimum.

Winter - 2022 MBAD 6141M - 084 30

30

15
 12/22/2021

Reorder Point and Safety Stock

31

When to Reorder: Reorder Point

 When the quantity on hand of an item drops to this amount,

the item is reordered.
o Determinants of the reorder point
1. The rate of demand
2. The lead time
3. The extent of demand and/or lead time variability
4. The degree of stock out risk acceptable to management
 Reorder Point: Under Certainty
ROP  d  LT
where
d  Demand rate (units per period, per day, per week)
LT  Lead time (in same time units as d )

Winter - 2022 MBAD 6141M - 084 32

32

16
 12/22/2021

Reorder Point: Under Uncertainty

 Demand or lead time uncertainty creates the possibility

that demand will be greater than available supply
 To reduce the likelihood of a stock out, it becomes
necessary to carry safety stock
o Safety stock
 Stock that is held in excess of expected demand due to variable
demand and/or lead time

Expected demand
ROP   Safety Stock
during lead time

Winter - 2022 MBAD 6141M - 084 33

33

Safety Stock

 As the amount of safety

stock carried increases, the
risk of stock out decreases.
o This improves customer
service level
 Service level
o The probability that demand
will not exceed supply
during lead time
o Service level = 100% - Stock
out risk

Winter - 2022 MBAD 6141M - 084 34

34

17
 12/22/2021

How Much Safety Stock?

 The amount of safety stock that is appropriate for a

given situation depends upon:
1. The average demand rate and average lead time
2. Demand and lead time variability
3. The desired service level
Expected demand
ROP   z dLT
during lead time
where
z  Number of standard deviations
 dLT  The standard deviation of lead time demand

Winter - 2022 MBAD 6141M - 084 35

35

Reorder Point: Demand Uncertainty

ROP  d  LT  z d LT
where
z  Number of standard deviations
d  Average demand per period (per day, per week)
 d  The stdev. of demand per period (same time units as d )
LT  Lead time (same time units as d )

Note: If only demand is variable, then  d LT   d LT

Winter - 2022 MBAD 6141M - 084 36

36

18
 12/22/2021

Reorder Point: Lead Time Uncertainty

ROP  d  LT  zd LT
where
z  Number of standard deviations
d  Demand per period (per day, per week)
 LT  The stddev. of lead time (same time units as d )
LT  Average lead time (same time units as d )

Note: If only lead time is variable, then  d LT  d  LT

Winter - 2022 MBAD 6141M - 084 37

37

Reorder Point: Both Demand and Lead

Time Have Uncertainty
ROP 𝑑 LT 𝑧𝑑𝜎LT
where
𝑧 Number of standard deviations
𝑑 Demand per period per day, per week
𝜎LT The stddev. of lead time same time units as 𝑑
LT Average lead time same time units as 𝑑

Note: If both lead time and __ __ 2

d  d LT    d 
2 2
demand are variable, then LT d LT

Winter - 2022 MBAD 6141M - 084 38

38

19
 12/22/2021

Example: Reorder Point

The housekeeping department of a motel uses approx. 400 washcloths per day.
The actual number tends to vary with the number of guests on any given night.
Usage can be approximated by a normal distribution that has a mean of 400
and std. dev. of nine washcloths per day. A linen supply company delivers
towels and washcloths with a lead time of three days. If the motel policy is to
maintain a stockout risk of 2%, what is the minimum number of washcloths that
must be on hand at reorder time, and how much of that amount can be
considered safety stock?
a. ROP for variable demand and constant lead time.
__
d = 400 washcloths per day Lead Time (LT) = 3 days
 d = 9 washcloths per day Risk = 2%, so service level = 98%  z = 2.055
_
ROP  d LT  z  d LT  400  3  2.055  9  3
 1200  32 .03  1232
Safety stock is approximately 32 washcloths
Winter - 2022 MBAD 6141M - 084 39

39

Example: Reorder Point (cont’d.)

The motel uses approx. 600 bars of soap each day, and this tends to be fairly
constant. Lead time for soap delivery is normally distributed with a mean of six
days and a std. dev. of two days. A service level of 90% is desired.
a. ROP for constant demand and variable lead time.
__
d = 600 bars per day Mean Lead Time, LT = 6 days
 LT = 2 days Service Level = 90%, so z = 1.28
___
ROP  d LT  z ( LT ) d  600  6  1.28  2  600
 3600  1536  5136 bars of soap
b. How many days of supply of soap?

ROP 5136
  8.56 days
d 600

Winter - 2022 MBAD 6141M - 084 40

40

20
 12/22/2021

Example: Reorder Point (cont’d.)

The motel replaces broken glasses at a rate of 25 per day. In the past, this
quantity has tended to vary normally and have a std. dev. of 3 glasses per day.
Glasses are ordered from a Cleveland supplier. Lead time is normally
distributed with an average of 10 days and a std. dev. of 2 days. What ROP
should be used to achieve a service level of 95%?

a. ROP for variable demand and variable lead time.

__ __
d = 25 glasses per day Mean Lead Time, LT = 10 days
 d = 3 glasses per day Service Level = 95%, so z = 1.65
 LT = 2 days
__ ___ __ __ 2
ROP  d LT  z LT  d2  d  2LT
2 2 2
 25  10  1.65 10 (3 )  ( 25 ) ( 2 )
 334 glasses

Winter - 2022 MBAD 6141M - 084 41

41

Fixed Order Interval Model

42

21
 12/22/2021

How Much to Order: FOI

 Fixed-order-interval (FOI) model

o Orders are placed at fixed time intervals
o Inventory brought up to target amount; amount ordered varies
 Reasons for using the FOI model
o Supplier’s policy may encourage its use
o Grouping orders from the same supplier can produce savings in shipping
costs
o Some circumstances do not lend themselves to continuously monitoring
inventory position
Expected demand
Amount Safety Amount on hand
during protection
to Order stock at reorder time
interval
𝑑 OI LT 𝑧𝜎 OI LT 𝐴
where
OI Order interval length of time between orders
𝐴 Amount on hand at reorder time

Winter - 2022 MBAD 6141M - 084 43

43

Fixed-Quantity vs. Fixed-Interval Ordering

Fixed Quantity

Fixed Interval

Winter - 2022 MBAD 6141M - 084 44

44

22
 12/22/2021

Example: Fixed Order Interval

A lab orders a number of chemicals from the same supplier every 30 days. Lead
time is 5 days. The manager of the lab must determine how much of one of these
chemicals to order. A check of stock revealed that eleven 25-miliiliter (ml) jars
are on hand; daily usage of the chemical is approx. normal with as mean of 15.2
ml per day and a std. dev of 1.6 ml per day. The desired service level for this
chemical is 95%.
a. How many jars of the chemical should be ordered?
__
d = 15.2 ml per day OI = 30 days SL = 95%  z = 1.65
 d = 1.6 ml per day LT = 5 days A = 11 jarsx25 ml per jar = 275 ml
__
Amount to Order  d (OI  LT )  z  d OI  LT  A
272 .62
 15.2(30  5)  1 .65 (1 .6 ) 30  5  275  272 .62 ml   10.9  11 jars
25
b. What is the average amount of safety stock of the chemical?
Safety Stock  z  d OI  LT  1 .65 (1 .6) 30  5  15 .62 ml
Winter - 2022 MBAD 6141M - 084 45

45

Single Period Model

46

23
 12/22/2021

Single Period Inventory Model

Consider the problem of deciding how

many newspapers to put in a hotel
lobby

Too few papers and some
customers will not be able to
purchase a paper, and profits
associated with these
potential sales are lost.
Too many papers and the
price paid for papers that
were not sold during the day
will be wasted, lowering
profit.

Winter - 2022 MBAD 6141M - 084 47

47

Single Period Model Applications

Overbooking of airline flights

Ordering of clothing and

other fashion items

One-time order for events –

e.g., t-shirts for a concert

Winter - 2022 MBAD 6141M - 084 48

48

24
 12/22/2021

Solving the Newspaper Problem

 Consider how much risk we are willing to take of running out of
inventory – that will be the service level or alternatively risk of not
being able to serve.
 Assume a mean of 90 papers and a standard deviation of 10 papers.
 Assume we want an 80 percent chance of not running out.
 Assume that the probability distribution associated of sales is normal,
stocking 90 papers yields a 50 percent chance of stocking out.

Winter - 2022 MBAD 6141M - 084 49

49

Solving the Newspaper Problem

 From Appendix B - Table B, we see that we need

approximately 0.85 standard deviation of extra
papers to be 80 percent sure of not stocking out.
 Using Excel, “=NORMSINV(0.8)” = 0.84162
 Stock 0.84162 x 10 = 8.4 extra papers or 9 papers

Winter - 2022 MBAD 6141M - 084 50

50

25
 12/22/2021

Single Period Model – Costs and Service

Levels
 Model for ordering of perishables and other items with limited useful
lives
 The goal of the single-period model is to identify the order quantity
that will minimize the long-run excess and shortage costs
 Two categories of problem:
o Demand can be characterized by a continuous distribution
o Demand can be characterized by a discrete distribution
o Shortage cost
 Generally, the unrealized profit per unit
 Cshortage = Cs = Revenue per unit – Cost per unit
o Excess cost
 Difference between purchase cost and salvage value of items left over at the end of
the period
 Cexcess = Ce = Cost per unit – Salvage value per unit

Winter - 2022 MBAD 6141M - 084 51

51

Stocking Levels

Cs
Service level 
C s  Ce
where
C s  shortage cost per unit
Ce  excess cost per unit
Cs Ce
Service level
Quantity
For a uniform distribution So So =Optimum
Balance Point Stocking Quantity

Winter - 2022 MBAD 6141M - 084 52

52

26
 12/22/2021

Example: Single Period Model

A Las Vegas supermarket bakery must decide how many wedding

cakes to prepare for the upcoming weekend. Cakes cost $33 each
to make and they sell for $60 each. Unsold cakes are sold at $10
each on Monday. Analysis of recent demand and probabilities
resulted in the following table: Demand Probability Cumulative
of Demand Probability
Shortage cost C S = $60 - $33 = $27 0 0.15 0.15
1 0.30 0.45
Cost of Excess C e = $33 - $10 = $23 2 0.25 0.70
3 0.20 0.90
Service Level = 0.54 4 0.10 1.00

Since 0.54 is between 0.45 and 0.70 cumulative probabilities,

stock at least 2 cakes to provide a service level of 0.54.

Winter - 2022 MBAD 6141M - 084 53

53

Example: Single Period Model

A firm that installs cable TV systems uses a certain piece of equipment for which
it carries two spare parts. The parts cost $500 each and have no salvage value.
Part failures can be modeled by a Poisson distribution with a mean of two failures
during the useful life of the equipment. Holding and disposal costs are
negligible. Estimate the apparent range of shortage cost.
Shortage cost C S = unknown; Cost of excess C e= $500
Number of Cumulative C S  0.406 Or
Failures Probability C S  0.406 (500  C S )
0 0.135
CS  Ce
1 0.406 C S  $341 .75
2 0.677
C S  0.677 Or
3 0.857 CS  0.677 (500  C S )
4 0.947 CS  Ce
5 0.983 C S  $1047 .99
… …. Range of shortage cost is $341.75 to $1047.99
Appendix B, Table C

Winter - 2022 MBAD 6141M - 084 54

54

27
 12/22/2021

PROJECT MANAGEMENT
Vinay Vasudev, Ph.D.

MBAD 6141M – 084 OPERATIONS MANAGEMENT

Projects
 Unique, one-time operations designed to accomplish a specific set
of objectives in a limited time frame
o The Olympic Games
o Producing a movie
o Product/Software development
o ERP implementation
 Nature of Projects
o Projects go through a series of stages – a life cycle
o Projects bring together people with a diversity of knowledge and skills, most of whom
remain associated with the project for less than its full life
o Organizational structure affects how projects are managed

Winter - 2022 MBAD 6141M - 084 2

1
 12/22/2021

Critical Project Roles

 Project Champion
o A person who promotes and supports a project and critical to
success
 Usually resides within the organization
 Facilitate the work of the project by ‘talking up’ the project to other
managers, and who might be asked to share resources with the project team
as well as employees who might be asked to work on parts of the project
 Project Manager
o The project manager is ultimately responsible for the
success or failure of the project
o The project manager must effectively manage:
 The work
 The human resources
 Communications
 Quality/Time/Costs/Performance – Project Management Triangle

Winter - 2022 MBAD 6141M - 084 3

Work Breakdown Structure (WBS)

 A hierarchical listing of what must be done during a project

o Establishes a logical framework for identifying the required activities for the
project
1. Identify the major elements of the project
2. Identify the major supporting activities for each of the major elements
3. Break down each major supporting activity into a list of the activities that will be needed to
accomplish it

Winter - 2022 MBAD 6141M - 084 4

2
 12/22/2021

Project Planning & Scheduling: Gantt Chart

Winter - 2022 MBAD 6141M - 084 5

Project Management Techniques: PERT

and CPM
 PERT (Program Evaluation and Review Technique) and CPM
(Critical Path Method) are two techniques used to manage
large-scale projects
 By using PERT or CPM Managers can obtain:
1. A graphical display of project activities
2. An estimate of how long the project will take
3. An indication of which activities are most critical to timely project
completion
4. An indication of how long any activity can be delayed without
delaying the project

Winter - 2022 MBAD 6141M - 084 6

3
 12/22/2021

Project Management Technique: Network

Diagram
 Network (Precedence) Diagram
o Diagram of project activities that shows sequential relationships (or
precedence) by use of arrows and nodes
o Activity on Arrow (AOA)
 Network diagram convention in which arrows designate activities
o Activity on Node (AON)
 Network convention in which nodes designate activities
 Activities - Project steps that consume resources and/or time
 Events - The starting and finishing of activities
o Time Estimates
 Deterministic (fairly certain)
 Probabilistic (allow variation)

Winter - 2022 MBAD 6141M - 084 7

Deterministic Computing Algorithm

 Finding ES (Early Start) and EF (Early Finish) involves a forward pass through
the network diagram
o Early Start (ES) – The earliest time an activity can start (Assumes that all activities
start as early as possible)
 ES = EF of the entering arrow
 ES = the largest of the entering EF
o Early Finish (EF) – The earliest time an activity can finish
 EF = ES + t, where t = activity time

 Finding LS and LF involves a backward pass through the network diagram

o Late Start (LS) – The latest time the activity can start and not delay the project
 LS = LF – t, where t = activity time
o Late Finish (LF) – The latest time the activity can finish and not delay the project
 LF = LS of the leaving arrow
 LF = the smallest of the leaving arrows

Winter - 2022 MBAD 6141M - 084 8

4
 12/22/2021

Slack and the Critical Path

 Slack is the allowable slippage for a path which can be computed one
of two ways:
o Slack = LS – ES
o Slack = LF – EF
 Critical path
o The longest path which determine the expected duration of the project
o The critical path is indicated by the activities (also, critical activities) with zero
slack
 Knowledge of slack times and critical paths provides managers with
information for planning allocation of scarce resources
 Control efforts will be directed toward those activities that might be
most susceptible to delaying the project

Winter - 2022 MBAD 6141M - 084 9

Example - CPM with Single Time Estimate

Consider the following new department setting-up project:

Activity Designation Immed. Pred. Time (Weeks)
Locate new facilities A None 8
Interview prospective staff B None 4
Hire and train staff C B 9
Select and order furniture D A 6
Remodel and install phones E A 11
Furniture received and set up F D 3
Move in/startup G C, E, F 1

Develop a critical path diagram (network) and determine the

duration of the critical path and slack times for all activities.

Winter - 2022 MBAD 6141M - 084 10

10

5
 12/22/2021

1. Draw the Network Diagram

Activity Immed. Pred. Time

A None 8
B None 4
C B 9
D(6)
D A 6
E A 11 F(3)
F D 3
A(8)
G C, E, F 1

E(11) G(1)
Start

B(4) C(9)

Winter - 2022 MBAD 6141M - 084 11

11

Find Out the Critical Path – Simple

Network
D(6)  A path is a sequence of
F(3)
activities that runs
A(8)
from the beginning to
E(11) G(1) the end of the project.
Start  The critical path is the
B(4) C(9) path where sum of the
activity times is the
Critical Path longest.

Path Length of the Path (wks)

A-D-F-G 8+6+3+1=18
A-E-G 8+11+1=20 *
Project Duration
B-C-G 4+9+1=14

Winter - 2022 MBAD 6141M - 084 12

12

6
 12/22/2021

2. Forward Pass: Compute ES and EF

Early Early
start finish 8 14
Activity D(6) 14 17
(duration)
0 8 F(3)
A(8)
8 19 19 20

E(11) G(1)
Start

0 4 4 13 Max{17, 19, 13}

B(4) C(9)

Winter - 2022 MBAD 6141M - 084 13

13

3. Backward Pass: Compute LS and LF

Early Early Min{8, 10}

start finish
8 14
Activity D(6) 14 17
(duration)
0 8 10 16 F(3)
Late Late
start finish A(8) 16 19
8 19 19 20
0 8
E(11) G(1)
Start 8 19 19 20

0 4 4 13

B(4) C(9)
6 10 10 19

Winter - 2022 MBAD 6141M - 084 14

14

7
 12/22/2021

4. Computing Slacks and Determining

Critical Path
ES EF
8 14
Activity D(6) 14 17
- (duration) 0 8 10 16 F(3)
A(8) 16 19
LS LF 8 19 19 20
0 8 E(11) G(1)
Start 8 19 20
19
0 4 4 13
B(4) C(9)
The critical Path: A-E-G
6 10 10 19

Activity LS-ES Slack On Critical Path

A 0-0 0 x
B 6-0 6
C 10-4 6
D 10-8 2
E 8-8 0 x

F 16-14 2
G 19-19 0 x

Winter - 2022 MBAD 6141M - 084 15

15

Critical and Non-critical Activities

 The activities on the critical path(s) (called critical activities)

have zero slack and cannot be delayed without delaying the
completion of the project.
 The slack for non-critical activities represents the amount of
time by which the start of these activities can be delayed
without delaying the completion of the entire project
(assuming that all predecessor activities start at their earliest
start times).

Winter - 2022 MBAD 6141M - 084 16

16

8
 12/22/2021

Probabilistic Time Estimates

 Time estimates that allow for variation

 The beta distribution is generally used to describe the inherent
variability in time estimates
 The probabilistic approach involves three time estimates:
o Optimistic time, (to)
 The length of time required under optimal conditions
o Pessimistic time, (tp)
 The length of time required under the worst conditions
o Most likely time, (tm)
 The most probable length of time required

Winter - 2022 MBAD 6141M - 084 17

17

Probabilistic Time Estimates

 The expected time, te ,for an activity is a weighted

average of the three time estimates:
to  4t m  t p
te 
6
 The expected duration of a path is equal to the sum
of the expected times of the activities on that path:

Path mean   of expected times of activities on the path

Winter - 2022 MBAD 6141M - 084 18

18

9
 12/22/2021

Probabilistic Time Estimates

 The standard deviation of each activity’s time is estimated as

one-sixth of the difference between the pessimistic and
optimistic time estimates. The variance is the square of the
standard deviation:
 t  t 
2

  p o 
2

 6 
 Standard deviation of the expected time for the path

 path   Variances of activities on path 

Winter - 2022 MBAD 6141M - 084 19

19

Knowledge of Path Statistics

 Knowledge of expected path times and their

standard deviations enables managers to compute
probabilistic estimates about project completion
such as:
o The probability that the project will be completed by a
certain time
o The probability that the project will take longer than
its expected completion time

Winter - 2022 MBAD 6141M - 084 20

20

10
 12/22/2021

Determining Path Probabilities

Specified time - Path mean

z
Path standard deviation

Winter - 2022 MBAD 6141M - 084 21

21

Project Completion Time

 A project is not complete until all project activities are

complete
 It is risky to only consider the critical path when assessing
the probability of completing a project within a specified
time.
o To determine the probability of completing the project within a
particular time frame
 Calculate the probability that each path in the project will be
completed within the specified time
 Multiply these probabilities
o The result is the probability that the project will be completed
within the specified time

Winter - 2022 MBAD 6141M - 084 22

22

11
 12/22/2021

Assumption: Independence

 Assumption is that path duration times are independent of each other

and requires that
1. Activity times are independent
2. Each activity is on only one path
 The assumption of independence is usually considered to be met if
only a few activities in a large project are on multiple paths
 When activity times cannot be assumed to be independent,
simulation is often used with repeated sampling
o Many paths are made through the project network
o In each pass, a random value for each activity time is selected based on the activity
time’s probability distribution
o After each pass, the project’s duration is determined
o After a large number of passes, there are enough data points to prepare a frequency
distribution of the project duration
o Probabilistic estimates of completion times are made based on this freq. dist.

Winter - 2022 MBAD 6141M - 084 23

23

CPM with Three Activity Time Estimates

 Procedures
Step 1: Calculate the expected time for each activity
Step 2: Determine the critical path.
Step 3: Calculate the variance of the activity times.
Step 4: Determine the probability of completing the
project on a given date.

Winter - 2022 MBAD 6141M - 084 24

24

12
 12/22/2021

Example – Three Time Estimates

Winter - 2022 MBAD 6141M - 084 25

25

Calculate the Expected Completion Time for a

Path and the Project
21 28 28 36
ET path   ETa C(7) F(8)

21 28 28 36
0 21 36 38

A(21) G(2)

0 21 36 38
21 26 26 28 28 33

B(5) D(2) E(5)

Critical Path (s) 21 26 26 28 31 36

Path 1: A-C-F-G 21+7+8+2=38

Path 2: A-C-E-G 21+7+5+2=35
Path 3: A-B-D-E-G 21+5+2+5+2=35 TE = Expected Completion Time for Project
Path 4: A-B-D-F-G 21+5+2+8+2=38

Winter - 2022 MBAD 6141M - 084 26

26

13
 12/22/2021

Calculate the Probability of Finishing the

Project in 35 weeks (or less)
 Calculate the variance of the expected time of the critical
path:

  CP
2
 Sum of the variances on the critical path

D  TE 35  38
z   0.87
 2
CP
11.89

Winter - 2022 MBAD 6141M - 084 27

27

Knowledge of Path Statistics

 Knowledge of expected path times and their

standard deviations enables managers to compute
probabilistic estimates about project completion
such as:
o The probability that the project will be completed by a
certain time
o The probability that the project will take longer than
its expected completion time

Winter - 2022 MBAD 6141M - 084 28

28

14
 12/22/2021

Determining Path Probabilities

Specified time - Path mean

z
Path standard deviation

Winter - 2022 MBAD 6141M - 084 29

29

Project Completion Time

 A project is not complete until all project activities are
complete
 It is risky to only consider the critical path when assessing the
probability of completing a project within a specified time.
o To determine the probability of completing the project within a
particular time frame
 Calculate the probability that each path in the project will be completed
within the specified time
 Multiply these probabilities
o The result is the probability that the project will be completed within
the specified time

Winter - 2022 MBAD 6141M - 084 30

30

15
 12/22/2021

Assumption: Independence
 Assumption is that path duration times are independent of each other and
requires that
1. Activity times are independent
2. Each activity is on only one path
 The assumption of independence is usually considered to be met if only a
few activities in a large project are on multiple paths
 When activity times cannot be assumed to be independent, simulation is
often used with repeated sampling
o Many paths are made through the project network
o In each pass, a random value for each activity time is selected based on the activity time’s
probability distribution
o After each pass, the project’s duration is determined
o After a large number of passes, there are enough data points to prepare a frequency
distribution of the project duration
o Probabilistic estimates of completion times are made based on this freq. dist.

Winter - 2022 MBAD 6141M - 084 31

31

Calculate the Probability from the z-Table

 Probability of finishing in 35 weeks (or less) is about

19% based on the critical path activities only

Winter - 2022 MBAD 6141M - 084 32

32

16
 12/22/2021

Budget Control

 Budget control is an important aspect of

project management
o Costs can exceed budget
 Overly optimistic time estimates
 Unforeseen events
o Unless corrective action is taken, serious
cost overruns can occur

Winter - 2022 MBAD 6141M - 084 33

33

Time-Cost Trade-Offs

 Activity time estimates are made for some given level of

resources
 It may be possible to reduce the duration of a project by
injecting additional resources
o To avoid late penalties
o Monetary incentives
o Free resources for use on other projects
 Crashing
o Shortening activity durations - typically, involves the use of additional
funds to support additional personnel or more efficient equipment, and
the relaxing of some work specifications
o The project duration may be shortened by increasing direct expenses,
thereby realizing savings in indirect project costs

Winter - 2022 MBAD 6141M - 084 34

34

17
 12/22/2021

Crashing Decisions

 To make decisions concerning crashing requires

information about:
1. Regular time and crash time estimates for each
activity
2. Regular cost and crash cost estimates for each activity
3. A list of activities that are on the critical path
o Critical path activities are potential candidates for crashing
o Crashing non-critical path activities would not have an impact
on overall project duration

Winter - 2022 MBAD 6141M - 084 35

35

Crashing: Procedure

 General procedure:
1. Crash the project one period at a time
2. Crash the least expensive activity that is on
the critical path
3. When there are multiple critical paths, find
the sum of crashing the least expensive
activity on each critical path
o If two or more critical paths share
common activities, compare the least
expensive cost of crashing a common
activity shared by critical paths with the
sum for the separate critical paths

Winter - 2022 MBAD 6141M - 084 36

36

18
 12/22/2021

Example: Project Crashing – Step 1

The indirect cost remains a constant ($10) for the first eight days and then increases at a rate of $5
per day if project takes longer than 8 days.
B(5)
$10  $6 $4
  $4
A(2) D(3) 2 1 1

C(4)
2-1=1

Normal Time Crash Normal Crash Cost per day to Number of days
(NT) Time Cost Cost expedite activity may be
Activity (NC) (CC) shortened (NT-
(CT)
CT)

A 2 1 $6 $10 $4 1
B 5 2 $9 $18 $3 3
C 4 3 $6 $8 $2 1
D 3 1 $5 $9 $2 2

Winter - 2022 MBAD 6141M - 084 37

37

Example : Project Crashing - Step 2

 Determine the critical path (using normal schedule):

B(5) Path Length

A(2) D(3) A-B-D 2+5+3=10

A-C-D 2+4+3=9
C(4)
Critical ppath
Activity direct cost = sum of all normal cost
= 6+9+6+5 = $26
Project indirect cost = 10 + 2*5 = $20
Total cost = 26 +20 = $46
Winter - 2022 MBAD 6141M - 084 38

38

19
 12/22/2021

Example: Project Crashing - Step 3 (1)

 Shorten project one day at a time. Check for change on the

critical path.
Critical Cost per day Available
B(5) Activity to crash days

A(2) D(3) A $4 1
B $3 3
C(4)
D $2 2 1
(1) Shorten activity D by one day.
Increases direct
B(5)
Path Length cost by $2 and
D(2) decreases indirect
A(2) A-B-D 9
cost by $5. Net
A-C-D 8 change is -$3 and
C(4) total cost is $43.

Winter - 2022 MBAD 6141M - 084 39

39

Example: Project Crashing - Step 3 (2)

 Continue to shorten project one day at a time. Check for change

on the critical path.
Critical Cost per day Available
B(5) Activity to crash days

A(2) D(2) A $4 1
B $3 3
C(4)
D $2 1 0
(2) Shorten activity D by one day.
Increases direct
B(5) cost by $2 and
Path Length
decreases indirect
A(2) D(1) A-B-D 8 cost by $5. Net
change is -$3 and
A-C-D 7
C(4) total cost is $40.

Winter - 2022 MBAD 6141M - 084 40

40

20
 12/22/2021

Example: Project Crashing - Step 3 (3)

 Continue to shorten project one day at a time. Check for change

on the critical path.
Critical Cost per day Available
B(5) Activity to crash days

A(2) D(1) A $4 1
B $3 3 2
C(4)
D $2 0
(3) Shorten activity B by one day.
Increases direct
B(4) cost by $3 and
Path Length DOES NOT
A(2) D(1) A-B-D 7 decrease the
indirect cost. Net
A-C-D 7 change is $3 and
C(4)
total cost is $43.
Winter - 2022 MBAD 6141M - 084 41

41

Example: Project Crashing - Step 3 (4)

 Continue to shorten project one day at a time. Check for change

on the critical path. Critical Cost per day Available
B(4) Activity to crash days
A $4 1 0
A(2) D(1)
B $3 2
C(4) C $2 1
D $2 0
(4) Shorten activity A by one day.
Increases direct cost
B(4) by $4 and DOES
Path Length
NOT decrease the
A(1) D(1) A-B-D 6 indirect cost. Net
A-C-D 6 change is $4 and
C(4) total cost is $47.

Winter - 2022 MBAD 6141M - 084 42

42

21
 12/22/2021

Example: Project Crashing - Step 3 (5)

 Continue to shorten project one day at a time. Check for change

on the critical path. Critical Cost per day Available
B(4) Activity to crash days
A $4 0
A(1) D(1)
B $3 2 1
C(4) C $2 1 0
D $2 0
(4) Shorten activities B and C by one day.
Increases direct cost
B(3)
Path Length by $5 and DOES
D(1) NOT decrease the
A(1) A-B-D 5
indirect cost. Net
A-C-D 5 change is $5 and
C(3) total cost is $52.

Winter - 2022 MBAD 6141M - 084 43

43

Example: Project Crashing - Step 3 (6)

Critical Cost per day Available

Activity to crash days
B(3)
A $4 0
A(1) D(1)
B $3 1
C(3) C $2 0
D $2 0

Cannot shorten the project anymore!

Winter - 2022 MBAD 6141M - 084 44

44

22
 12/22/2021

Example: Project Crashing - Step 4

Plot of Costs and Minimum-Cost Schedule

Minimum

Winter - 2022 MBAD 6141M - 084 45

45

Example: Project Crashing - Summary

 The crashing sequence is summarized as below:

LENGTH AFTER CRASHING n DAYS

Path n=0 1 2 3 4 5

A-B-D 10 9 8 7 6 5

A-C-D 9 8 7 7 6 5

Activity crashed D D B A B, C

Direct Cost 26 28 30 33 37 42

Indirect Cost 20 15 10 10 10 10

Total Cost 46 43 40 43 47 52

Winter - 2022 MBAD 6141M - 084 46

46

23
 12/22/2021

Project Risk Management

 Risks are an inherent part of project management

o Risks relate to occurrence of events that have undesirable
consequences such as
 Delays
 Increased costs
 Inability to meet technical specifications
o Good risk management involves
 Identifying as many risks as possible
 Analyzing and assessing those risks
 Working to minimize the probability of their occurrence
 Establishing contingency plans and budgets for dealing with any
that do occur

Winter - 2022 MBAD 6141M - 084 47

47

24
 3/12/2022

MANAGEMENT OF WAITING LINES

Vinay Vasudev, Ph.D.

MBAD 6141M - 084 OPERATIONS MANAGEMENT

Waiting Lines

 Waiting lines occur in all sorts of service systems

 Wait time is non-value added
o Wait time ranges from the acceptable to the emergent
 Short waits in a drive-thru
 Sitting in an airport waiting for a delayed flight
 Waiting for emergency service personnel
o Waiting time costs
 Lower productivity
 Reduced competitiveness
 Wasted resources
 Diminished quality of life

Winter - 2022 MBAD 6141M - 084 2

1
 3/12/2022

Queuing Theory

 Mathematical approach to the analysis of waiting

lines
o Applicable to many environments
 Call centers
 Banks
 Post offices
 Restaurants
 Theme parks
 Telecommunications systems
 Traffic management

Winter - 2022 MBAD 6141M - 084 3

Why Is There Waiting?

 Waiting lines tend to form even when a system is not fully
loaded
o Variability
 Arrival and service rates are variable
o Services cannot be completed ahead of time and stored for later
use
 Waiting Lines – Managerial Implications:
1. The cost of providing waiting space
2. A possible loss of business when customers leave the
line before being served (balking) or refuse to wait at
all
3. A possible loss of goodwill
4. A possible reduction in customer satisfaction
5. Resulting congestion may disrupt other business
operations and/or customers
Winter - 2022 MBAD 6141M - 084 4

2
 3/12/2022

Waiting Line Management

 The goal of waiting line management is to minimize total costs:

 Costs associated with customers waiting for service
 Capacity cost

Winter - 2022 MBAD 6141M - 084 5

Waiting Line Characteristics

 The basic characteristics of waiting lines

1. Population source
2. Number of servers (channels)
3. Arrival and service patterns
4. Queue discipline
System

Processing Order
Calling
population Arrivals Waiting Service Exit
line

Winter - 2022 MBAD 6141M - 084 6

3
 3/12/2022

Waiting Line Characteristics

 Population Source
o Infinite source
 Customer arrivals are unrestricted

 The number of potential customers greatly exceeds system capacity
o Finite source
 The number of potential customers is limited
 Channel
o A server in a service system
o It is assumed that each channel can handle one customer at a time
 Phases
o The number of steps in a queuing system

Winter - 2022 MBAD 6141M - 084 7

Common Queuing Systems

Winter - 2022 MBAD 6141M - 084 8

4
 3/12/2022

Waiting Line Characteristics

 Arrival pattern
o Most commonly used models assume the arrival rate can be described by
the Poisson distribution - Arrivals per unit of time
o Equivalently, interarrival times are assumed to follow the negative
Exponential distribution - The time between arrivals
 Service pattern
o Service times are frequently assumed to follow a negative exponential
distribution
 Queue discipline - The order in which customers are processed
o Most commonly encountered rule is that service is provided on a first-
come, first-served (FCFS) basis
o Non FCFS applications do not treat all customer waiting costs as the same

Winter - 2022 MBAD 6141M - 084 9

Waiting Line Metrics/Performance

 Managers typically consider five

measures when evaluating waiting
line performance:
1. The average number of customers
waiting (in line or in the system)
2. The average time customers wait (in
line or in the system)
3. System utilization
4. The implied cost of a given level of
capacity and its related waiting line The average number waiting in line and
the average time customers wait in line
5. The probability that an arrival will increase exponentially as the system
utilization increases
have to wait for service

Winter - 2022 MBAD 6141M - 084 10

5
 3/12/2022

Queuing Models: Infinite Source

 Four basic infinite source models
o All assume a Poisson arrival rate
1. Single server, exponential service time
2. Single server, constant service time
3. Multiple servers, exponential service time
4. Multiple priority service, exponential service time
Infinite Source Symbols
  Customer arrival rate Ws  Average time customers spend in the system
  Service rate per server 1   Service time
Lq  Average number of customers waiting for service P0  Probability of zero units in the system
Ls  Average number of customer in the system
Pn  Probability of n units in the system
r  Average number of customers being served
  System utilization M  Number of servers (channels)
Wq  Average time customers wait in line Lmax  Maximum expected number waiting in line

Winter - 2022 MBAD 6141M - 084 11

Basic Relationships

 System Utilization
𝜆
𝜌
𝑀𝜇
 Average number of customers being served
𝜆
𝑟
𝜇
 Little’s Law
o For a stable system the average number of customers in line or in the system
is equal to the average customer arrival rate multiplied by the average time
in the line or system
𝐿 𝜆𝑊
𝐿 𝜆𝑊

Winter - 2022 MBAD 6141M - 084 12

6
 3/12/2022

Basic Relationships

 The average number of customers

o Waiting in line for service:
Lq [Model dependent.]
o In the system:
𝐿 𝐿 𝑟

 The average time customers are

𝐿
o Waiting in line for service 𝑊
𝜆
o In the system 1 𝐿
𝑊 𝑊
𝜇 𝜆
Winter - 2022 MBAD 6141M - 084 13

Single Server, Exponential Service Time

 M/M/1
𝜆
𝐿
𝜇 𝜇 𝜆
𝜆
𝑃 1
𝜇
𝜆
𝑃 𝑃
𝜇
𝜆
𝑃 1
𝜇

Winter - 2022 MBAD 6141M - 084 14

7
 3/12/2022

Example 1 (page 792): Infinite Source

Single Server Model – M/M/1
Customers arrive at a bakery at an average rate of 18 per hour on
weekday mornings. The arrival distribution can be described by a
Poisson distribution with a mean of 18. Each server can serve a
customer in an average of three minutes; this time can be described
by an Exponential distribution with a mean of 3.0 minutes.
a. What are the arrival and service rates?
The arrival rate =  = 18 customers/hour
The service rate = = 3 minutes/customer = 20 customers/hour

b. Compute the average number of customers being served at

any time. 𝜆 18
𝑟 0.90 customers
𝜇 20
Winter - 2022 MBAD 6141M - 084 15

Example 1: Infinite Source Single Server

Model – M/M/1 (cont’d.)
c. Suppose it has been determined that the average number of
customers waiting in the line is 8.1. Compute the average
number of customers in the system (i.e., waiting in line or
being served), the average time customers wait in line, and the
average time in systems.
Avg. number of customer waiting in line = L q = 8.1 customers

Avg. number of customer waiting in system = L s  L q  r  8 .1  0 . 9  9 . 0

L q  8.1  0.45 hour
Avg. time customer is in the line = W q 
 18
1 1
Avg. time customer is in the system = W s  W q   0.45   0.50 hour
 20

Winter - 2022 MBAD 6141M - 084 16

8
 3/12/2022

Example 1: Infinite Source Single Server

Model – M/M/1 (cont’d.)
d. Determine the system utilization for M=1, 2, 3 servers.

System Utilization = 
M

18
For M = 1,   0.90
1 20

18
For M =2,   0.45
2 20
18
For M =3,   0.30
3 20
Winter - 2022 MBAD 6141M - 084 17

Example 2 (page 794): Infinite Source

Single Server Model – M/M/1
An airline is planning to open a satellite ticket desk in a new
shopping plaza, staffed by one ticket agent. It is estimated that
requests for tickets and information will average 15 per hour, and
request will have a Poisson distribution. Service time is assumed to
be exponentially distributed. Previous experience with similar
satellite operations suggests that mean service time should
average about three minutes per request.
  15 customers per hour
1 customer
  60  20 customers per hour
3 minutes
 15
a. System Utilization =    0.75
M 1  20
Winter - 2022 MBAD 6141M - 084 18

9
 3/12/2022

Example 2: Infinite Source Single Server

Model – M/M/1 (cont’d.)
b. Percentage of time the server (agent) will be idle.
% of idle time = 1    1  0.75  0.25  25 %

c. Expected number of customers waiting to be served.


2 2
15
Lq   (   )

20 ( 20  15)
 2.25 customers

d. The average time customers will spend in the system.

L q 1 2.25 1
W s      15  20  0.20 hour or 12 minutes

Winter - 2022 MBAD 6141M - 084 19

Example 2: Infinite Source Single Server

Model – M/M/1 (cont’d.)
e. The probability of zero customers in the system and the
probability of four customers in the system

 15
P 0  1    1  20  0.25

4
𝜆 15 4
𝑃4 𝑃0 0.25 = 0.079
µ 20

Winter - 2022 MBAD 6141M - 084 20

10
 3/12/2022

Example 2: Infinite Source Single Server

Model – M/M/1 (cont’d.)
e. The probability of zero customers in the system and the
probability of four customers in the system

 15
P 0  1    1  20  0.25

2 4
  15 
P 4  P 0     0.25 20   0.079
   

Winter - 2022 MBAD 6141M - 084 21

Single Server, Constant Service Time

 M/D/1
o If a system can reduce variability, it can shorten
waiting lines noticeably
o For, example, by making service time constant, the
average number of customers waiting in line can be
cut in half 2
Lq 
2 (   )
o Average time customers spend waiting in line is also
cut by half.
o Similar improvements can be made by smoothing
arrival rates (such as by use of appointments)
Winter - 2022 MBAD 6141M - 084 22

11
 3/12/2022

Example 3 (page 794): M/D/1

Wanda’s Car Wash & Dry is an automatic, five-minute operation

with a single bay. On a typical Saturday morning, cars arrive at a
mean rate of eight per hour, with arrivals tending to follow a
Poisson distribution.
  8 cars per hour
1 car
  60  12 cars per hour
5 minutes
a. Avg. number of cars in line =

2 2
8
L q  2  (    )  2  12 (12  8)  0.667 cars
b. Avg. time in the system =
L q 1 0.667 1
W s      8  12  0.167 hour or 10 minutes

Winter - 2022 MBAD 6141M - 084 23

Multiple Servers

 Assumptions: 
M

  
– A Poisson arrival rate and   
Lq  P
exponential service time M  1! M    2 0
1
– Servers all work at the  
n

M

 M 1      
same average rate 
P0         
– Customers form a single  n  0 n!   
 M !  1   
waiting line (in order to   M   
maintain FCFS 1
Ws 
processing) M  
W
PW  q
Ws

Winter - 2022 MBAD 6141M - 084 24

12
 3/12/2022

Cost Analysis

 Service system design reflects the desire of management to

balance the cost of capacity with the expected cost of
customers waiting in the system
 Optimal capacity is one that minimizes the sum of customer
waiting costs and capacity or server costs

Winter - 2022 MBAD 6141M - 084 25

Maximum Line Length

 An issue that often arises in service system design is how much

space should be allocated for waiting lines
 The approximate line length, Lmax, that will not be exceeded a
specified percentage of the time can be determined using the
following:
log K ln K
L max  or
log  ln 
where
specified
1 
percentage
K 
L q 1   

Winter - 2022 MBAD 6141M - 084 26

13
 3/12/2022

Problem 6 (page 817): M/M/2

Trucks are required to pass through a weighing station so that

they can be checked for weight violations. Trucks arrive at the
station at the rate of 40 an hour between 7:00 pm and 9:00 pm.
Currently two inspectors are on duty during those hours, each of
whom can inspect 25 trucks an hour.
  40 trucks per hour   25 trucks per hour M 2
a. How many trucks at the station (system)?
 40
L s  L q    2.844  25  4.444 trucks
b. How long driver will expect to be in the system?
L q 1 2.844 1
W s      40  25  0.1111 hour or 6.666 minutes

Winter - 2022 MBAD 6141M - 084 27

Problem 6: M/M/2 (cont’d.)

c. Prob. both inspectors are busy at the same time.
L q 2.844
W q    40  0.0711
W q 0.0711
Pw  
0.10
 0.7111
Wa
d. How many minutes on average a truck will wait on arrival not
immediately served?
1 1 1
W a  M    (50  10 )  10 hour  6 minutes
e. Only one inspector  line would grow
f. What is the max line length for prob. = 0.97?
 40 1  0.97 ln k
   0.80 k  0.05274 L max  ln   13 .186
M 2  25 2.844 (1  0.80 )
Winter - 2022 MBAD 6141M - 084 28

14
 3/12/2022

Psychology of Waiting

 If those waiting in line have nothing else to occupy their

thoughts, they often tend to focus on the fact that they are
waiting in line
o They will usually perceive the waiting time to be longer than the actual
waiting time
o Steps can be taken to make waiting more acceptable to customers
 Occupy them while they wait
 In-flight snack
 Have them fill out forms while they wait
 Make the waiting environment more comfortable
 Provide customers information concerning their wait

Winter - 2022 MBAD 6141M - 084 29

15