Review for Exam II

1
Review for Exam II

• This exam will be administered
• Friday,July 22, 2011, beginning at 1000 a.m.

2
Exam Format

• 40-50 multiple choice
• 3 problems
• Closed-book
• Closed-notes
• Closed-neighbor
• BRING---pencil, calculator, orange scantron sheet

3
Exam details

• Will be administered on Friday,July 22, 2011
• Will start the exam at 1000 a.m.

4
Exam Coverage

• Chapter 10, Chapter 11, Supplement to Ch 11,
  Chapter 13, Supplement to Ch14, and Chapter
  15-second half
• LP problems in the supplement to Chapter 14, but
  not the content of Chapter 14we will cover that
  later.

5
Typical problemssee Practice Exam II

• Inventory with Independent Demand
• Multiple Choice Problems involving calculation of
  inventory turns and days of supply
• Production Scheduling Problem
• Solved using ROW MINIMUM METHOD (RMM)
• Transportation problem
• Solved using RMM

6
More typical problems

• Transshipment problem
• Linear programming formulation
• Be able to draw schematics of mainframe/glass
  architecture, client/server architecture and
  N-tier architecture

7
Chapter 15 ERP

• Wont be tested on Inventory for Dependent
  Demand, just the material on ERP, beginning on
  page 700 and continuing through 709

8
What were five motivations for transitioning from
mainframes

• Absence of data integration
• 36 month backlogs at centralized MIS shops
• Idle CPU cycles on desktops
• Mainframes were expensive bottlenecks
• Support for Internet and thin clients
• Quicker, cheaper development times through REUSE

9
What is the information architecture modern ERP
systems are currently based on?

• Mainframe/glass house
• Client/server
• N-tier distributed
• None of these

10
Every application software package consists of

• Presentation management component
• Business logic management component
• Data management component
• All of the above
• 1 and 2 only

11
ERP

• Is software that organizes and manages a
  companys business processes by sharing info
  across functional areas
• Large caps have been there and done that
• Mid and small caps are getting there
• The road to implementation has been rough

12
More ERP

• Based on an N-tier distributed architecture
• Not on mainframe glasshouse

13
Advantages of N-tier architecture

• Provides for data integration
• Better usage of MIPS on both PCs and servers
• Solves the 36-month backlog of the centralized
  MIS shop
• Enables a better career path for the MIS
  professional

14
N-Tier architecture

• Is decentralized or centralized, or some
  combination of these (which?)
• Utilizes thick clients or thin clients (which?)

15
ERP Modules

• Sales distribution
• Production Materials Management
• Quality management
• Human resource management

• Project management
• Accounting and controlling/finance
• Supply chain management
• Customer relationship management

16
ERP Terms

• Best-of-breed
• Collaborative product commerce
• Customer relationship management
• Supply chain management
• XML

17
Re-engineered Computer Architectures

• Started with mainframe/glasshouse
• Migrated to client/server
• Evolved to N-tier distributed

18
Why did such re-engineering occur?

• There was no data integration
• MIPs on mainframes were hugely expensive and very
  much in demand
• MIPs on PCs were idle 95 of the time and
  extremely cheap
• Backlogs for MIS shops were at 36 months
• Developing new applications were slow and
  expensive

19
Distributed architectures solved these problems

• Data resides behind a single database engine

20
Components of any Software Application
Presentation Management
Business Logic
Data Management
Database
21
Components in brief
PM
BL
DM
22
Mainframe Architecture (circa 1993)
Mainframe Computer
PM
PM
PM
BL
BL
BL
DM
DM
DM
23
Problems with Mainframe Architecture

• Absence of data integration, resulting in little
  enterprise visibility
• The applications are maintainable only by the
  centralized MIS shop, which is overloaded,
  resulting in 36 month lead times to get revisions
  effected
• Every application had to be built from scratch,
  line-by-line, resulting in large cost and long
  lead times to create new applications

24
More problems with Mainframe Architecture

• No reuse was possible
• These mainframe apps were accessed on networked
  PCs via IBM 3278 terminal emulation software
  that was completely incompatible with the windows
  GUI applicationsmeaning no cut and paste
• Mainframes were computational bottlenecks
• Desktop PCs sat idle 99 of the time

25
First solution Client/server architecture
Clients (PM, BL)
Server (DM)
Database
26
These were known as thick clients

• Because they contained both the presentation
  management (PM) and the business logic (BL)
  components of the application
• Notice how the application is distributed across
  the network, residing in two computing boxesthe
  client or desktop and the server

27
First solution Client/server architecture
Thick Clients (PM, BL)
Server (DM)
Database
28
Advantages of Client/server architecture

• All Data are all accessible behind the Server
  which runs the data management portion of the
  applicationusually an Oracle Database engine
• Now the marketing guy can see where his
  customers job is, and whether the customer is
  current with his payments, among other things

29
Advantages of client/sever architecture

• The IT professional could sit shoulder-to-shoulder
  with the end-user and develop applications as
  well as make changes to existing software
  rapidly, without a 36 month backlog
• For new applications, there were huge reuse
  opportunitiesin particular, the IT professional
  does not have to create a DM componentthe Oracle
  engine can be reused

30
Problems with Client/server

• It wasnt Internet compatible
• It required an IT professional to install
  software on the end-users personal computer (the
  client)
• It required an IT professional to work closely
  with the non-IT professional
• There were no career paths for IT professional
  hired in marketing, finance, accounting,
  manufacturing, etc.

31
Modern solution of today N-TIER DISTRIBUTED

• This is a distributed architecture like
  client/server, but now the application is
  distributed across three or more computing boxes
  on the network

32
N-Tier distributed Architecture
Thin Clients (1/2PM)
Data Server (DM)
Database
Application Server 1
Application Server 2
33
Take a closer look at the Application Servers
Application Server runs the business logic
component and half ot the presentation management
componentthe portion the serves out the web pages
34
Comments on N-Tier Distributed Architecture

• Clients are called thin because the only thing
  running on them is the Internet Browser
• The IT professional doesnt have to install
  anything on the client
• More re-use is possiblespecifically that browser

35
Advantages of N-Tier Distributed Architecture

• Like Client/server, it accommodates enterprise
  visibility because the data are integrated
• Applications can be built rapidly because there
  is abundant reuse
• The DM module is reused
• Half of the PM component is reused
• There are reuse opportunities within the rest of
  the PM component and the BL component as well

36
More advantages of N-Tier

• IT professionals dont have to be remotely loaned
  out to marketing, management, accounting and
  finance
• They can now be centrally located and managed
  where career paths will exist for them

37
Application Servers do Two things

• They serve out web pages upon request
• They do all of the business logic processing.

38
ERP Modules

• Finance/Accounting
• Sales Marketing
• Production/Materials Management
• Human Resources

39
These modules would be placed in a

• Thin client
• Data server
• Application server
• Mainframe

40
ERP Implementation

• Analyze business processes
• Choose modules to implement
• Align level of sophistication
• Finalize delivery and access
• Link with External Partners

41
Customer Relationship Management

• CRM software plans and executes business
  processes that involve customer interaction, such
  as marketing, sales, fulfillment, and service
  (not manufacturing)
• CRM is focused on customers, not products

42
Collaborative Product Commerce

• Software concerned with new product design and
  development, as well as product lifecycle
  management

43
Connectivity

• A common data management component
• APIs (Application Programming Interfaces)
• EAI (Enterprise Application Integration)
• XML (Extensible Markup Language)
• Dr. Viator (accounting) teaches a course in this
  language

44
Chapter 10--Supply Chain Management

• Plants/warehouses/distribution/information
  infrastructure
• Most of Americas product gets moved by _____
  (air, water, rail, truck, pipeline).
• What is COVISINT??
• What benefits accrue from SCM?

45
Whats new and exciting in SCM??

• Information Technology (specifically enterprise
  visibility)
• Has changed everything
• SCM Software modules within ERP systems
• I2 Technologies
• Has reduced uncertainty
• Which has reduced _____________
• Which is a form of _______________

46
Inventory turns

• Calculated on an annual basis
• The more, the better
• Inputs
• Cost of goods sold
• Average aggregate value of inventory

47
Average aggregate value of inventory

• Calculated by taking the product of the unit cost
  with the number of units and then summing these
  products for all inventory categories

48
Days of supply

• Avg agg value of Inv365/Ann cost of goods sold
• Or simply 365/inventory turns

49
Manufacturing Inventory Types

• Raw materials inventory
• Work-in-process inventory
• Finished goods inventory

50
Supply Chain Management Terms

• Bullwhip effect
• Collaborative planning, forecasting and
  replenishment
• Continuous replenishment
• Core competencies
• Cross-docking
• E-business
• E-marketplaces
• E-procurement
• EDI
• Inventory turns

• Landed cost
• Logistics
• Order fulfillment
• RFID
• Sourcing
• Vendor-management inventory
• Warehouse management system

51
Chapter 13 Inventory Management

• Inventory for Independent demand
• Not manufacturing inventory, usuallymore like
  retail inventory

52
Carrying costs

• Rent
• Lighting/heating
• Security
• Interest (on borrowed capital tied up in
  inventory)
• Taxes
• Shrink/obsolescence/theft
• Can also be expressed as a of product cost
• A rule of thumb is 30

53
Ordering costscosts related to

• Transportation
• Shipping
• Receiving
• Inspection

54
Shortage costs

• This is an opportunity cost
• Is ignored in the simple models you will be
  using, by assuming that there are no shortages

55
Back-order costs

• Will assume impatient customers who must have the
  product they wish to buy NOW.
• So back-ordering is not considered in the simple
  models we looked at

56
Continuous Inventory Systems

• Constant order amount, called the EOQ
• EOQ Economic Order Quantity
• Fixed annual deterministic demand
• Minimizes
• Holding (carrying) costs
• Ordering costs
• Uses re-order point to determine when to order
• Time between orders is not fixed

57
EOQ models also have

• No shortages/back-ordering
• Constant lead time
• Instantaneous or finite replenishment
• Can take into consideration price discounting
• When doing so, three costs are minimized jointly
  Ordering costs, holding costs and purchase costs
  taken over a years time

58
If the quantity ordered is less than the EOQ, then

• Ordering costs will be greater than holding
  (carrying) costs

59
ABC Classificationwhat is the point??

• To concentrate, focus on the those items in
  inventory that constitute the highest dollar
  value to the firm
• Class A items constitute 5-15 of the items and
  70 to 80 of the total dollar value to the firm
• Class B items constitute 30 of the inventory
  items but only 15 of the dollar value
• Class C items constitute 50 to 60 of the items
  but only 5 to 10 of the dollar value

60
ABC Classification..

• Class A items are tightly controlled
• Class B items less so
• Class C items even less
• Dollar values are computed by multiplying the
  unit cost by the annual demand for the item
• This technique is used in all auto parts
  inventory control systems and has been for 15
  years

61
Periodic inventory systems are.

• Fixed Time period systems
• NOT
• EOQ Models
• The time between orders is fixed, the re-order
  point is fixed, but the order amount is not

62
Which gives you lowest holding cost?

• Instantaneous replenishment
• Finite (non-instantaneous) replenishment
• Quantity discounts
• WHICH OF THE ABOVE GIVES YOU LOWEST TOTAL
  ORDERING COST?

63
How do we calculate a re-order point?

• Lead time in days times the daily demand plus the
  safety stock
• Safety stock equals the service level (usually 3
  for z) times the standard deviation of daily
  demand times the sq. rt. of lead time.
• (You will be given the formulas)

64
How do we calculate

• Time between orders?
• Production days in a year / of orders
• Run length
• EOQ or order quantity / daily Production rate

65
Safety Stocks and Service Levels

• Safety stock Z value std. dev. of daily
  demand sqrt(lead time)
• For 95 service level, use Z value of 1.65
• For 99 service level, use Z value of 3

66
Inventory Terms

• Independent demand
• Inventory
• In-process inventory
• Non-instantaneous receipt
• Order cycle
• Quantity discount
• Stockout
• Service level
• Efficiency

• ABC system
• Carrying costs
• Continuous inventory system
• Dependent demand
• EOQ
• Fixed-order quantity system
• Fixed time period system
• Capacity

67
Simulation

• Two types
• Continuous deterministic
• VENSIM is an example
• Discrete stochastic
• PROMODEL is an example
• Each of these two types differ by method of time
  advance

68
Time advance in continuous deterministic
simulation

• Time is advanced in small, equidistant increments
• The simulation engine is really integrating
  differential equations

69
Time advance in discrete stochastic simulation

• Time is advanced from event to event
• The simulation engine maintains a stack of
  discrete events chronologically ordered in time,
  called an events calendar
• The next event to occur is popped off the stack
  and processed.
• The result of processing the event is that more
  events are generated and subsequently get saved
  on the events calendar

70
MONTE CARLO

• The computer-generation of random numbers using an

Algorithm
71
Which simulation gestalt uses activities, events,
entities and their attributes?

• Continuous deterministic?
• Discrete stochastic?

72
The Excel function RAND() generates

• Normally-distributed random variates
• Gamma-distributed random variates
• Uniformly-distributed random numbers
• Exponentially-distributed random variates

73
To get a non-uniform random variate, we often
start with

• A normal random variate
• A lognormal random variate
• A uniform random number
• A triangular random variate

74
To get a non-uniform random variate, we often use

• The central limit theorem
• The law of large numbers
• The inverse function theorem
• All of the above

75
In discrete/stochastic simulation, we are
interested in

• Entity idleness
• Entity travel time
• Entity time in the system
• Resource utilization
• All of the above

76
In discrete/stochastic simulation, which of the
following components has time duration?

• Events
• Activities
• Entities
• Resources
• All of the above

77
Discrete/stochastic simulation is appropriate for
which of the following three decision environments

• Decision Making (DM) under Certainty
• DM under risk and uncertainty
• DM under change and complexity

78
Math programming models, like the transportation
and transshipment models we looked at, are
appropriate for which decision making environment

• Decision Making (DM) under Certainty
• DM under risk and uncertainty
• DM under change and complexity