Investigación Operativa

Operations Research

https://rossetti.github.io/RossettiArenaBook/

Introduction to
ARENA

Departamento de Organización Industrial

September 1, 2024
 1. Example: The Pharmacist
2. Definition of variables and how to obtain
statistics on variables
3. Extended Example: The Pharmacist
Continued

Example 1: The Pharmacist

https://www.rockwellautomation.com/es-es/products/software/arena-simulation.html
 A demo example
• Customers arrive at a drive-through
pharmacy window according to a
Poisson distribution with a mean of 10
customers per hour.
• The time it takes the pharmacist to
serve the customer is random, and
data indicate that the time is well
modeled with an exponential
distribution with a mean of 3 minutes.
• Customers who arrive at the
pharmacy are served in the order of
arrival, and we assume that enough
space is available within the parking File: Pharmacist_chapter_1.doe
area to accommodate any waiting
customers.

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 3

Operations Research/Investigación Operativa
 Arena Environment

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 4

Operations Research/Investigación Operativa
 Processing Entities: Basic Process/Data Modules (i)
Basic Process templates:
• CREATE: To create entities that pass through the model
• DISPOSE: To dispose of the entities
• PROCESS: Set of activities experienced by an entity
‐ Delay (Delay entity for the service/activity; no recourse needed)
‐ Seize Delay Release (Seize resource; delay for the service/activity;
release resource)
• DECIDE: Direct the entity through the system by probabilistic and
conditional mechanisms (if/else; while).

Basic Data templates:

• Entity: It is possible to create specific entities (not just generic)
• Resource: Define the resources of your problem.
• Variable/Attribute: Define state variables or Define entity attributes
• Queue: Queue of entities that form in front of processes.

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 5

Operations Research/Investigación Operativa
 Creating and Disposing of Entities in ARENA
• CREATE module: the way to introduce entities into a model
Time Between Arrivals
Types:

Random: Uses an Arrival rate as a

Exponential distribution Variable
Schedule: Nonhomogenous Compound arrival:
Poisson process with rates
specified by Schedule module Disc(0.2,1, 0.5, 2, 1.0,3)

Constant: Constant time 20% arrival of 1

30% arrival of 2
50% arrival of 3
Expression: List of distributions
or expressions:
Normal, Triangular, Beta, …
Entities per Arrival: Number of entities that will
enter the system at a given time with each arrival

Max Arrivals: Maximum number of arrival events

allowed by the module. It is helpful to limit the
number of entities flowing through the model.

First Creation: Starting time for the first entry

to arrive at the system (not applied when the Type is Schedule)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 6

Operations Research/Investigación Operativa
 Creating and Disposing of Entities in ARENA
• DISPOSE module: the entities that come into this module leave the model

Time Statistics: Cost Statistics:

Value Added Time (VA Time) Value Added Cost (VA Cost)
Non-Value Added Time (NVA Time) Non-Value Added Cost (NVA Cost)
Wait Time Wait Cost
Total Time Total Cost,…

Run > Setup > Project Parameters

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 7

Operations Research/Investigación Operativa
 Processing Entities: Basic/Advanced Process Modules (ii)

• SEPARATE / BATCH: To duplicate or combine entities in a model

• ASSIGN: To change/update/initialize the value of various attributes and
variables
• READWRITE / FILES: To handle inputs and outputs in a model
• RECORD / STATISTIC: To collect statistics in a module

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 8

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (i)
• Using the Basic Process Panel Template, drag the CREATE, PROCESS, and DISPOSE
into the model window and connect them via the module connector.

Module
connector
Drag and
drop from
Basic
process
panel

CREATE PROCESS DISPOSE

module module module

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 9

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (ii)
• Arrival Process
– CREATE module: Poisson process with an arrival rate of 10 customers per hour (6
minutes between two consecutive arrivals)
– Double-click on the icon to show its module panel and include parameters

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 10

Operations Research/Investigación Operativa
 Relation between Poisson and Exponential

Poisson distribution Exponential distribution

Arrival rate of Interarrival time of
10 customers per hour 6min

 = 10 cl/60min = 1/6 1/  = 6

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 11

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (iii)
• ENTITY module
– ENTITY: Elements that flow through the system (pharmacist’s customers)
– Add a new entry using the double-click in the spreadsheet window
– Associate a picture to the entity

• RESOURCE module
– An element demanded by entities with finite capacity
– Edit the new entry by clicking in the spreadsheet window

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 12

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (iv)
• PROCESS module
– PROCESS: Set of activities
experienced by an entity
– Processes may be resource-
unconstrained or constrained
(example)
– Pharmacist’s example:
• Seize the pharmacist Choose Seize Delay Release
• Delay in the service
• Release the pharmacist
Click Add to get Resources usage
dialog

Choose
expression

Exponential service times

with a mean of 3

This is the number of units of the resource needed by the

entity (not the resource capacity)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 13

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (v)
• RUN Setup
– Assumptions:
• The pharmacy is open 24 hours a day, 7 days a week
• The arrival process does not vary over time
• Simulation time: 10,000 hours of operation
• Go to the Run menu item and choose Setup
• The base time units are minutes for output reports
• To run the model:
– Run menu item
– Run toolbar
Fast Start over
Run forward

Step Pause Stop

Fast forward:
Run without animation (Batch mode)
Start over:
Stops the run and starts it again
Animation slider:
To slow down or speed up
Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 14
Operations Research/Investigación Operativa
 Implementing the demo model in Arena (v)
• RUN Setup

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 15

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (vi)
• ANALYZE Results
– EXCEL Pharmacist chapter_1.rpt

– Sheet “Across replication summary”

Discrete-Time Statistics (Tally)

Average StDev Of Minimum Maximum Overall Overall Avg Observations

Name Type Source Of Replication Averages Observed=Requested Half-Width Replication Averages Replication Average Replication Average Min Value Max Value Per Replication
customer NVA Time Entity 0 1 0 0 0 0 0 0 100350
Other Time Entity 0 1 0 0 0 0 0 0 100350
Total Time Entity 6,12271116 1 0,144155099 0 6,12271116 6,12271116 2,01E-05 70,525213 100350
Transfer Time Entity 0 1 0 0 0 0 0 0 100350
VA Time Entity 3,030109759 1 0,021750103 0 3,030109759 3,030109759 2,01E-05 37,451817 100350
Wait Time Entity 3,092601401 1 0,127711816 0 3,092601401 3,092601401 0 68,422888 100350
Get Medicine.Queue Waiting Time Queue 3,092570583 1 0,127711816 0 3,092570583 3,092570583 0 68,422888 100351

Half width of the

Total number of
confidence interval
customers
Continuous-Time Statistics (Time Persistent)
Average Of StDev Of Min Max Overall
Replication Replication Replication Replication Overall Max
Project Name Name Type Source Averages Half-Width Averages Average Average Min Value Value
Unnamed Project Get Medicine.Queue Number Waiting Queue 0,517237584 0,018575505 0 0,51723758 0,517237584 0 16
Pharmacist Instantaneous Utilization Resource 0,506787198 0,004501153 0 0,5067872 0,506787198 0 1
Number Busy Resource 0,506787198 0,004501153 0 0,5067872 0,506787198 0 1
Number Scheduled Resource 1 2E+20 0 1 1 1 1
customer WIP Entity 1,024024782 0,022177505 0 1,02402478 1,024024782 0 17

51% busy

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 16

Operations Research/Investigación Operativa
 Implementing the demo model in Arena (vii)
• ANALYZE Results (cont.)
– Text-based report: the file “modelname.out” in the directory associated to the model

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 17

Operations Research/Investigación Operativa
 1. Example: The Pharmacist
2. Definition of variables/attributes and how
to obtain statistics on variables
3. Extended Example: The Pharmacist
Continued

Definition of variables/attributes and

how to obtain statistics on variables
 Defining Variables and Attributes
• VARIABLES: quantities that are properties of the system that may change
• Suggestion: its name starts with v (for example, vLambda)
• Variables are used to share information across all modules in the model
• To define a variable is possible through the spreadsheet view or through a dialog box

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 19

Operations Research/Investigación Operativa
 Defining Variables and Attributes (cont.)
• ATTRIBUTE: It is a named property or characteristic of an entity
•Suggestion: its name starts with my (for example, myArrivalTime)
•For example, dimensions, color, reference, etc.
• There are predefined attributes that ARENA uses to track entities
- IDENT: to identify entities (each entity has a different value)
- Entity.SerialNumber: this number is unique when the entity is created,
but it is the same if the entity is cloned
- Entity.Type: It is useful to distinguish entities after they are created. Also,
it is used for the entity picture and cost attributes

Predefined attributes User-defined attributes

IDENT Entity.SerialNumber Entity. Type Size Weight Processing Time

1 1001 1 2 33 20
2 1002 2 3 22 55
3 1003 2 1 11 44
4 1004 1 2 33 20
5 1005 1 5 10 14
6 1006 3 4 14 10

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 20

Operations Research/Investigación Operativa
 Example in Arena
How to obtain variable of interest A) in ARENA

How to model How to see the results

Time-based • Create the variable • Mark “report statistics” on creation

• Update the value with ASSIGN • Then view statistics in the output
variable
through the process report (time persistent)

• Use RECORD to assign definite value for

Observation-based • Create the attribute associated the entity
variable with the entity • Then view statistics in the output
• Use ASSIGN to give initial value to report (section Tally)
the attribute

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 21

Operations Research/Investigación Operativa
 Example in Arena
How to obtain variable of interest A) in ARENA

To obtain the average WIP of rings

between F and G process points:
Time-based variable
vNumInMakeAndInspect

G
F

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 22

Operations Research/Investigación Operativa
 Example in Arena
How to obtain variable of interest A) in ARENA

How to collect
statistics on a
Time-based
variable?

The average number of

pairs of rings in both the
ring-making and the ring-
inspection process

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 23

Operations Research/Investigación Operativa
 Example in Arena
How to obtain variable of interest B) in ARENA

To collect observation-based statistics, it is

necessary to place RECORD modules at
appropriate locations in the model.

When the entity passes the

point F, there is an assignment: F
myArriveToMakeRingTime =TNOW

When the entity passes the RECORD module (of Type: Time Interval):
The module records the difference between the current time, TNOW, and the specified
attribute (myArriveToMakeRingTime)

This represents the elapsed time interval for the entity to move from F to the RECORD.

After sampling the required number of

replications of the new average time
obtained follows:
Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 24
Operations Research/Investigación Operativa
 Example in Arena
How to obtain variables of interest in ARENA

There are five RECORD types:

Count: A counter is defined, and it will be

incremented or decremented by the value specified

Entity Statistics: This option will record entity statistics

Time Interval: The difference between the current time,

TNOW, and the specified attribute

Time Between: The difference between the current time, TNOW, and the last time an
entity passed through this RECORD module

Expression: It allows statistics to be collected on a general expression when the entity

passes the module

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 25

Operations Research/Investigación Operativa
 1. Example: The Pharmacist
2. Definition of variables and how to obtain
statistics on variables
3. Extended Example: The Pharmacist
Continued

Extended Example:
The Pharmacist Continued
 Extending/modifying the demo example

• Compound Arrivals: 20% 1 customer, 30% 2 c., and 50% 3 c.

• Customers who arrive at the pharmacy have 1, 2, or 3

prescriptions with 50%, 30%, and 20% probabilities

• The service time mean is 3, 4, and 5 minutes, respectively

• Statistics to collect are the average number of prescriptions in

the system and the average time spent by the customer

• Collecting only the first 30 arrivals

File: Pharmacist_chapter2.doe

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 27

Operations Research/Investigación Operativa
 System Performance of the demo model
Variables:
• vNumPrescriptions: Scalar variable that counts the total number of prescriptions in
the system (initial v.=0)
• vNP: a 1D variable with 3 rows and counts the number of customers having 1, 2,
and 3 prescriptions currently in the system
• vServiceMean: a 1D variable with 3 rows representing the mean of the service
distributions for the 3 prescription quantities (initial values = 3, 4, and 5)

In order to collect
time-based statistics
Click here to enter
initial values
Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 28
Operations Research/Investigación Operativa
 Implementing the demo model in ARENA
• Arrival Process
– CREATE module: with a defined distribution of the number of customers
– We limit to 30 arrivals (cars, not customers)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 29

Operations Research/Investigación Operativa
 ASSIGN module (where values of variables/attributes are assigned)

DISC(cp1,v1,cp2,v2,…)
Discrete probability function
with probabilities:

𝑃(𝑋 = 𝑣 ) = 𝑐𝑝 − 𝑐𝑝

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 30

Operations Research/Investigación Operativa
 ASSIGN module (where values of variables/attributes are assigned)

In this module:
• We assign to this entity an attribute: the number of prescriptions (1, 2, or 3) with a
probability
• We add to the counter the overall number of prescriptions in the system and the
number of prescriptions associated with this entity
• We also assign to this entity the current simulation time (NOW) as his initial time in
the system
• We specify this is an arrival event (type 1)
• Finally, we compute in the vector vNP (clients with X prescriptions) in row “myNP” the
counter +1
Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 31
Operations Research/Investigación Operativa
 PROCESS module

This way to index arrays is very

common

In this module:
• We adapt the process module for getting the medicine with a delay time
depending on the number of prescriptions of the entity
• We go to the vector “vServiceMean” and read the “myNP” value

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 32

Operations Research/Investigación Operativa
 ASSIGN module (where values of variables/attributes are assigned)

In this module:
• We decrease the variable of the overall number of prescriptions, the number of
prescriptions of the entity that is leaving the system (“myNP”)
• We also decrease 1 from the overall counter of the number of clients with NP
prescriptions in the corresponding position
• We specify this is a leaving event (type 2)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 33

Operations Research/Investigación Operativa
 READWRITE module (read from/write to a file)
• The READWRITE module helps to debug the model (analyzing arrival and departure
times) in the Advanced Process Panel
• First, it is necessary to define a FILE in the Advanced Process Panel

Sequential File: A text file whose records will be in the same sequence as
they are written
Free format: each value in the file is separated by a space

Written Information
TNOW: Current Simulation Time
vEventType: Type of Event
IDENT: Entity identity number
Entity.SerialNumber
myNP: number of prescriptions
for the customer

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 34

Operations Research/Investigación Operativa
 READWRITE to FILE

FILE: Customer Trace.txt

2.076914 1.000000 2.000000 1.000000 1.000000
2.076914 2.000000 4.000000 1.000000 1.000000
2.498945 1.000000 2.000000 2.000000 1.000000 0.422032 2.076914
6.400697 3.000000 3.000000 1.000000 3.000000
6.400697 4.000000 6.000000 1.000000 1.000000
6.400697 5.000000 7.000000 1.000000 3.000000
11.064971 2.000000 4.000000 2.000000 1.000000 8.988057 2.076914
26.578563 3.000000 3.000000 2.000000 3.000000 20.177866 6.400697
26.941828 4.000000 6.000000 2.000000 1.000000 20.541131 6.400697
30.096515 5.000000 7.000000 2.000000 3.000000 23.695817 6.400697
37.113203 6.000000 5.000000 1.000000 1.000000

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 35

Operations Research/Investigación Operativa
 RECORD module
• Module in the Basic Process Panel
• This module “records” information
each time an entity passes
through it
• It is necessary to compute the
average time in the system
• The options included are:
– Count: Increase or decrease the value
of the named counter by a specified
value (useful to obtain statistics about
how many entities flow through the
model)
– Entity Statistics: Generate general
entity statistics (time and cost/duration
information)
– Time Interval: To calculate and record
the difference between a specified
attribute value and the current
simulation time
– Time Between: To track and record the
time between entities entering the
module
– Expression: To record the value of the
specified expression

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 36

Operations Research/Investigación Operativa
 Results
Two ways (you can choose in Run set up/Reports/Default report)

Model statistics SIMAN summary report

EXCEL txt in the app

Sheet: “AcrossReplicationsSummary” • Record System Time: time in the system

• customer.VATime: average time in the queue
Blocks • customer.WaitTime: average time in queue
• Discrete-Time Statistics (Tally) • vNumPrescriptions average number of
• Continuous-Time Statistics (Time-Persistent) prescriptions in the system
• Counter Statistics (Reports End of Replication • Pharmacist.Utilization: %occupancy of the resource
Value) • Get Medicine.Queue.NumberInQueue: average
• Output Statistics (Reports End of Replication Value) number in queue
• customer.NumberIn: total customers entering the
system
• customer.NumberOut: total customers entering the
system

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 37

Operations Research/Investigación Operativa
 Run the model with 30 first arrivals

If you entered a value,

clear the field and press to
get infinite
Maximum number of arrivals
(cars, not customers)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 38

Operations Research/Investigación Operativa
 Run the model with 30 first arrivals and 100 replications

Maximum number of arrivals

(cars, not customers)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 39

Operations Research/Investigación Operativa
 Half width when running 100 replications
Average Of Minimum Maximum
Replication Replicatio Replication Overall Min Overall Max
Name Type Source Averages Half-Width n Average Average Value Value
customer NVA Time Entity 0 0 0 0 0 0
Other Time Entity 0 0 0 0 0 0
Total Time Entity 49,649867 4,582213169 11,05208 113,50657 0,00127955 187,6067421
Transfer Time Entity 0 0 0 0 0 0
VA Time Entity 3,7236943 0,095717817 2,714779 4,9221801 0,00120285 35,620971
Wait Time Entity 45,926173 4,518547816 8,254803 109,22431 0 182,5127451
Get Medicine.Queue Waiting Time Queue 45,926173 4,518547816 8,254803 109,22431 0 182,5127451
User
Record System Time Expression Specified 49,649867 4,582213169 11,05208 113,50657 0,00127955 187,6067421

Average Of Min Max

Replication Replication Replication Overall Min Overall Max
Name Type Averages Half-Width Average Average Value Value
Get Medicine.Queue Number Waiting 11,47456726 1,02515029 2,532649887 25,18568559 0 46
Instantaneous
Pharmacist Utilization 0,944840599 0,011075687 0,754883768 0,999498194 0 1
Number Busy 0,944840599 0,011075687 0,754883768 0,999498194 0 1
Number Scheduled 1 0 1 1 1 1
customer WIP 12,41940786 1,030863922 3,309110876 26,17311828 0 47
vNumPrescriptions Variable 21,21781674 1,815009862 5,686477252 46,05937476 0 79

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 40

Operations Research/Investigación Operativa
 Augment the simulation:
Displaying variable values
• Display the number of prescriptions in the System (vNumPrescriptions)

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 41

Operations Research/Investigación Operativa
 Augment the simulation:
Plot the Current Number in Queue along the Simulation

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 42

Operations Research/Investigación Operativa
 Augment the simulation:
Plot a histogram of the pharmacist current utilization

Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 43

Operations Research/Investigación Operativa
 ARENA Typical Commands – Distributions/Other commands
Examples of distributions
Exponential: EXPO(value)
Triangular: TRIA(minimum, value, maximum)
Normal: NORM(mean, standard deviation)
Uniform: UNIF(minimum, maximum)
Other useful stuff
TNOW: Current simulation time. TNOW records the simulation clock time as the model
progresses.
TFIN: Final simulation time. TFIN is the ending time scheduled for the replication; it is a
real-valued quantity.
DISC: Discrete cumulative distribution function. DISC(cumulative probability 1, value 1,
cumulative prob.1, value 2, …, c.p.N, value N).
Commands for resources
MR(Resource ID): Resource capacity. MR returns the number of capacity units currently
defined for the specified Resource ID.
NR(Resource ID): Number of busy resource units.
ResUtil(Resource ID): Resource utilization. ResUtil returns the instantaneous utilization of
a resource as a real number between 0 and 1, inclusive.
Check out the ARENA Variables Guide
https://cours.etsmtl.ca/gpa662/documents/Cours/Arena%20Variables%20Guide.pdf
Grado en Ingeniería en Tecnologías Industriales/de Telecomunicación Discrete Event Simulation Modeling 44
Operations Research/Investigación Operativa
 Pedro Sánchez ([email protected])
Andrés Ramos ([email protected])
Eugenio Sánchez ([email protected])
Sonja Wogrin([email protected])
Gonzalo Alarcón ([email protected])

www.comillas.edu

www.comillas.edu