Simulation Modeling: A Business Process Analysis Tool

1
Simulation Modeling A Business Process Analysis
Tool

• BA 295 Fall 2002
• Prof. Fred Coleman, III
• October 8, 2002

2
SimulationWhat Is It?

• Technique for using computers to imitate or
  simulate the operations of various kinds of
  real-world facilities or processes
• -Law and Kelton, 3rd Edn, 2000

3
SimulationWhy Use It?

• Some systems simple enough to evaluate
  analyticallytherefore we use mathematical
  methods
• Algebra, calculus, probability theory
• Most Real-world systems too complex for
  analytical models
• We use simulation to evaluate a model of the
  real-world numerically and we estimate true
  characteristics.

4
SimulationWhat Does It Require for Business
Processes

• Process Map of business process
• Relationships/linkages
• Time for process interaction
• Cost for process interaction
• Perspective of item moving through system
• Data quality
• Above
• Appropriate probability distribution for time to
  handle item or item waits

5
Simulation

• The objective is to model the processes and
  activities of any system at a desired level of
  accuracy and realism balanced against the cost to
  obtain input data.
• Cost for input data is acquisition, scope,
  cleaning, formatting, modeling for input, delays,
  etc.

6
Three Dimensions to Classify Simulation Models

• Static
• Time plays no role in the analysis.
• Example Toss of a die a few hundred times.
• Estimates number of each die face
• Estimate is not correct

• Dynamic
• Time does play a role.
• Modeling any real system where variables and
  parameters are related to time.
• Production over time

7
Three Dimensions to Classify Simulation Models

• Continuous
• State of the system can change continuously over
  time.
• Precipitation
• Water flowing over a dam
• Flow in a pipeline

• Discrete
• Change can occur only at separated points in
  time. (i.e Event times)
• Bank teller model has events that occur only when
  customer arrives or customer leaves

8
Three Dimensions to Classify Simulation Models

• Deterministic
• Models have no random input
• Fixed service times like an appointment book.

• Stochastic
• Operate with random input
• Random arrivals to a fast food drive through
  window
• Waiting time for delivery of food

9
Combinations within Simulation Models

• Mixed Continuous and Discrete
• Refinery with continuously changing pressure and
  discretely occurring shutdowns
• Deterministic and Random Inputs
• Some inputs deterministic 7am11pm grocery store
  operating time (auto locks synch with NIST Atomic
  clock)
• Some inputs random time to be processed in
  grocery check-out

10
Entities

• Entities are the items of interest moving through
  system
• Electronic representations of customers,
  documents, parts
• Business Processes checks, contracts, purchase
  orders, etc.
• Service Systems people banks, retail,
  restaurants, etc.
• Mfg raw materials, subcomponents, finished
  product, etc.

11
Processes

• In general, these are the activity areas where
  entities are worked on or stored or delayed.
• Resource (machine or person) Queues (storage)
• Also these are the flow mechanisms that act on
  how the entity is moved or transported through a
  system.
• The simulation software demo tonight will refer
  to these as modules

12
Modules

• Create makes the entity and determines their
  rate of creation
• Process identifies the resources to act on the
  entity and how long they will operate
• Decide allows for a binary type decision
  yes/no, good/bad, true/false and how many of each
• Dispose exits the system, provides basic count
  of total(s)