when should i use simulation?
DESCRIPTION
When should I use simulation? Choosing the right process improvement tool for your project. Learn how an experienced engineer decides when simulation is the right tool for his projects, and when it isn't. With the evolution of process improvement software, it can be difficult to decide the right tool for the job. Using something too powerful and complex can be a lengthy and unnecessary process, but underestimating the depth of analysis required and choosing something too simplistic early in a project can result in repeated work later.TRANSCRIPT
When should I use
simulation?
Prof. Brian Harrington
SIMUL8 Corporation | SIMUL8.com | [email protected]
Agenda
• Common Manufacturing issues
• Intro to different types of simulation
• Using maths to analyze a Queuing System
• Using Excel/Monte Carlo simulation
• Using Discrete Event Simulation to look at
system design
• Six Sigma simulations
• A case study.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Manufacturing Dilemma
• Any product development process
involves extensive prototyping;
• Yet, costly manufacturing production
systems are typically not prototyped
SIMUL8 Corporation | SIMUL8.com | [email protected]
Simulation in Manufacturing
• System Design
• Operational Procedures
• Performance Evaluation
SIMUL8 Corporation | SIMUL8.com | [email protected]
System Design
• Plant Layout
• Effects of introducing new equipment
• Location and sizing of inventory buffers
• Location of inspection stations
• Optimal number of carriers, pallets
• Resource planning
• Protective capacity planning
Biggest Bang for the Dollar!
Contains Operational Procedures &
Performance Metrics.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Operational Procedures
• Production Scheduling - Choice of scheduling
and dispatching rules
• Control strategies for material handling
equipment
• Shift patterns and planned downtime
• Impact of product variety and mix
• Inventory Analysis
• Preventative maintenance on equipment
availability
Continuous Improvement
SIMUL8 Corporation | SIMUL8.com | [email protected]
Performance Evaluation
• Throughput Analysis (capacity of the
system, identification of bottlenecks); Jobs
per Hour
• Time-in-System Analysis
• Assessment of Work-in-process (WIP)
levels
• Setting performance measure standards;
OEE
If you can measure it, you can manage it!
SIMUL8 Corporation | SIMUL8.com | [email protected]
Agenda
• Common Manufacturing issues
• Intro to different types of simulation
• Using maths to analyze a Queuing System
• Using Excel/Monte Carlo simulation
• Using Discrete Event Simulation to look at
system design
• Six Sigma simulations
• A case study.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Why Simulation?
• Competition drives the following:
• Leaner production environment
• Shorter product development cycles
• Narrower profit margins
• Flexible Manufacturing (1 Facility, 1
Process, Multiple Models)
SIMUL8 Corporation | SIMUL8.com | [email protected]
Types of Simulation
• Mathematical Modeling
– e.g. Queuing Theory
• Monte Carlo Simulation
– e.g. Excel based models
• Discrete Event Simulation
– e.g. SIMUL8
SIMUL8 Corporation | SIMUL8.com | [email protected]
Simulation Overview
System Model
Deterministic Stochastic
Static Dynamic Static Dynamic
Continuous Continuous Discrete Discrete
DES
Monte Carlo
Differential equations
Queuing Theory
SIMUL8 Corporation | SIMUL8.com | [email protected]
Agenda
• Common Manufacturing issues
• Intro to different types of simulation
• Using maths to analyze a Queuing System
• Using Excel/Monte Carlo simulation
• Using Discrete Event Simulation to look at
system design
• Six Sigma simulations
• A case study.
SIMUL8 Corporation | SIMUL8.com | [email protected]
A Queuing System
Jockeying
Queue
Queue
Reneging
Service
Mechanism
Queue Structure Service Process
Arrival
Process
Balking
Serv
ed
Cu
sto
mers
Input Source
SIMUL8 Corporation | SIMUL8.com | [email protected]
Queuing Concepts Relationships for M/M/C
P = o
1
S n=0
C-1 (l/m) n
n!
c + (l/m)
c! ( )
cm
cm - l
L = q
(l/m)
2
c (l m) o P
(c – 1)! (cm – l)
l = mean arrival rate
m= mean service rate
C = number of parallel servers
These are messy to calculate by
hand, but are very easy with
appropriate software or a table.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Queuing Concepts A Comparison of Single Server Models
L = q
2(1 - l/m)
2 l s + (l/m)
2 2
L = q
2(1 - l/m)
2 (l/m)
L = q
(1 - l/m)
(l/m) 2
M/G/1
M/D/1
M/M/1
Note that
M/D/1 is
½ of M/M/1
SIMUL8 Corporation | SIMUL8.com | [email protected]
Limitations on Queuing Models
• What if:
– we don’t have one of these basic models?
– we have a complex system that has segments
of these basic models and has other
segments that do not conform to these basic
models?
• Then – simulate!
SIMUL8 Corporation | SIMUL8.com | [email protected]
Excel Based Simulations
• Uses Data Table functions
• Each Row might be one iteration of a simulation
• Each Col is a random variable generated in the
simulation
• RAND(), VLOOKUP(), COUNTIF(), NORMINV()
• Calculation & Iteration
• >>> Using VBA to bring in Probability functions
SIMUL8 Corporation | SIMUL8.com | [email protected]
Monte Carlo Simulation
• Named after the gaming tables of Monte Carlo
• Also referred to as a Static Simulation Model in
that it is a representation of a system at a
particular point in time
• In contrast, a Dynamic Simulation is a
representation of a system as it evolves over
time
• Might be accomplished using Excel and the
Random()
SIMUL8 Corporation | SIMUL8.com | [email protected]
Monte Carlo Simulation A Simple Example
Day RN Deman
d
Units
Sold
Units
Unsold
Units
Short
Sale
s
Rev
Return
s
Rev
Unit
Cost
Good
Will
Profit
$
1 10 16 16 2 0 4.80 0.16 2.70 0.00 2.26
2 22 16 16 2 0 4.80 0.16 2.70 0.00 2.26
3 24 17 17 1 0 5.10 0.08 2.70 0.00 2.48
4 42 17 17 1 0 5.10 0.08 2.70 0.00 2.48
5 37 17 17 1 0 5.10 0.08 2.70 0.00 2.48
6 77 18 18 0 0 5.40 0.00 2.70 0.00 2.70
7 99 20 18 0 2 5.40 0.00 2.70 0.14 2.56
8 96 20 18 0 2 5.40 0.00 2.70 0.14 2.56
9 89 19 18 0 1 5.40 0.00 2.70 0.07 2.63
10 85 19 18 0 1 5.40 0.00 2.70 0.07 2.63
Avg 2.50
Where do this numbers come from?
SIMUL8 Corporation | SIMUL8.com | [email protected]
Limitations & Disadvantages
• Stochastic, but static! Usually the time
evolution of a manufacturing system is
significant!
• Excel based models, soon start to use
VBA, and become very complicated
• Might require 1000’s of iterations; Data
Tables become slow
• Difficult to communicate results to
management.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Agenda
• Common Manufacturing issues
• Intro to different types of simulation
• Using maths to analyze a Queuing System
• Using Excel/Monte Carlo simulation
• Using Discrete Event Simulation to look at
system design
• Six Sigma simulations
• A case study.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Benefits of using DES Simulation
• Mathematical & Excel based models only go so
far
• Less difficult than mathematical methods
• Adds lot of “realism” to the model. Easy to
communicate to end users and decision makers
• Time compression
• Easy to “scale” the system and study the effects
• User involvement results in a sense of
“ownership” and facilitates implementation
SIMUL8 Corporation | SIMUL8.com | [email protected]
SIMUL8 Common Building Blocks
The 8 Common Building Blocks: Start Point, Queue, Activity, Conveyor,
Resource, and End Point. Then the Logical aspect Labels & Conditional
Statements.
SIMUL8 Corporation | SIMUL8.com | [email protected]
8 is all you Need
1. Work Item Types: Can represent parts,
carriers, signals, phone calls, just about
anything that requires a “Label Profile”.
2. Activities: Work Centers, machines, tasks,
process steps, anything that requires a “Cycle
Time”.
3. Storage Areas: Buffers, de-couplers, banks,
magazines, anything that requires a finite space
to occupy over time.
4. Conveyors: Moving parts from pt A to pt B;
Number of parts & Speed of conveyor.
SIMUL8 Corporation | SIMUL8.com | [email protected]
…8 is all you Need…
5. Resources: Manpower, crews, forklifts, tugs;
anything that require a certain resource to be
present.
6. End Pt: Keep track of statistics and free
memory!
7. Labels: The attributes of a Work Item.
8. Visual Logic: The ability to create conditional
statements; variables, loops, commands &
functions.
SIMUL8 Corporation | SIMUL8.com | [email protected]
Less is More using 6-Sigma
DMAIC or DMADV steps: • Define, Measure, Analyze, Improve, Control
• Define, Measure, Analyze, Design, Verify
DES Steps: • Objective, Assumptions, Data Collection, Build Model,
Verify, Validate, Experimentation, Results
Very similar steps!