“large and complex organizations are a tangible manifestation of advanced technology ......

University of Pisa

MSc in Computer Engineering

Business Processes Management

“Large and complex organizations are a tangible manifestation of advanced technology, more than machinery itself.” (J.K. Galbraith)

Lectures

Mario G. Cimino, Department of Information Engineering, Center for Logistics Systems

Pisa, March-May 2013, Monday 15.30-18.30, Room: ADInform1

Cimino – Modeling and Simulation of Business Processes using BPMN 2.0 – Lecture 1 – 2 of 15

Course Schedule and Materials

� The course will take about 30 hours in total:

2/3 of the time (March-April) for learning notation and tools (in parallel)

1/3 of the time (May) for working on the group project

� Materials:

http://www.iet.unipi.it/m.cimino/bpm

user: business

password: pr0cess (note: 0 is a zero)

� References:

OMG BPMN 2.0 Specification:

http://www.omg.org/spec/BPMN/2.0/PDF/

http://www.omg.org/spec/BPMN/2.0/examples/PDF

T. Allweyer, Introduction to the Standard for Business Process Modeling

2009, http://bpmn-introduction.com

Logizian 10.x Simulacian, a business workflow design and simulation

tool, http://www.visual-paradigm.com/product/lz/


Business Process Model and Notation (BPMN 2.0)

• BPMN is…

� human-readable: a standard visual notation for modeling business

processes;

� accessible: easy to understand for various roles: who analyzes and

defines processes, who leads the technological implementation, who is

responsible for management and control;

� machine-readable: a notation serializable to XML for process execution

(e.g. WS-BPEL 2, SOA environments).

• BPMN is not…

� a language for representing data flows and object flows, although this

can be done at a certain abstraction level;

� a notation to represent structures, functional decompositions, data

models, organization strategies, business rules.


• “Relatives” of BPMN (languages for worlfkow-based analysis)

� Petri Nets (1962): formal language to model distributed systems, usable

by computer scientist and designers of specialized software. It consists of

a visual representation and a corresponding mathematical notation

(graphs), which allow advanced analyses such as validation, verification

(e.g.. soundness to identify deadlock, livelock, ...)

� UML Activity Diagram (OMG, 1997): language for visual modeling for

the object-oriented paradigm, usable by software engineers. The extended

UML of Eriksson e Penker (2000) is suitable for business process

modeling, and usable also by business level (non-technical) roles.


Generic example of a process diagram, with UML extension of Eriksson-Penker

e.g. drilling

process of a

metal plate


• BPMN (OMG, 2005) is specialized in representing the behavior of processes

concerning the control flow, with the concept of a token traversing the process

structure.

• A Start Event1 generate a token that will be consumed by an End Event

2. The path of

tokens is managed by a network of Sequence Flow3, Gateway

4, Activity5 and

Intermediate Event6, within the process.

• Race pattern: there is a race between two intermediate events6 after the event-based

gateway4, i.e., “receive acknowledgment” and “four days elapsed”. When the

message is not received before the 4 days cutoff date, the execution is diverted from

normal execution flow in order to raise a ‘Partner Failure’ error (throw semantic)7.

http://www.iet.unipi.it/m.cimino/bpm/res/mov01.swf

3

1 2 4

5

6

6

6

2,7

5


• Interruption pattern: the previous model is reused and embedded as the ‘Reporting’

sub-process (expanded notation)1. An intermediate event (‘catch’ semantic) has been

added to the boundary of that sub-process. The model contains also a collapsed sub-

process (details are not visible), denoted by a “plus” sign4.

• The ’Partner failure’ event (‘catch’ semantic) gets activated when the execution

points reaches the ‘Reporting’ sub-process2, and gets un-activated when the

‘Reporting’ sub-process completes successfully (i.e., an end event is reached).


1

2

3

4


• Note how the partner failure event (‘send’ semantic) raised by the reporting sub-

process is caught by the ‘Partner failure’ event3 (‘receive’ semantic) and how the

execution flow get diverted from that point.

• Note that if a sub-process of a diagram is expanded in the diagram, the elements

inside the sub-process cannot be connected to elements outside the sub-process.

• BPMN allows the structured modeling of processes, i.e., views at different levels of

abstraction: from the level “0” (the least amount of detail) the processes are

decomposed into sub-processes, up to activities (that are atomic, the most amount of

detail). For instance, in the IMB methodology the analysis stops at the level “3”.



• At the level 0 process is a high-level not executable process, i.e., which has been

modeled for the purpose of documenting process behavior at a modeler-defined level

of detail. Thus, information needed for execution, such as formal condition

expressions are typically not included. In contrast, an executable process is modeled

for the purpose of being executed (e.g. a WS-BPEL process).

• BPMN allows process segmentation, at a given level, to create different modular

segments. For instance, in the IBM methodology it is recommended that the

maximum number of process activities per page should be six.

• The off-page connector, generally used for printing, is an object showing where a

Sequence Flow leaves one page and then restarts on the next page. A Link

Intermediate Event with throw1 and catch2 semantic can be used as an Off-Page

Connector. In figure, the flow of the exclusive gateway3 labeled with “No” leads to

come back, and then to a cycle.

1 2

3


• As a result of a ‘Sequential Loop’ marker1 on the Sub-Process, the sub-process will

be instantiated several times sequentially. As a result of a Parallel Loop’ marker2 on

the Sub-Process, the sub-process will be instantiated several times in parallel.

• The number of loops to execute might be: (i) defined at design time, (ii) affected at

runtime from some process data, (iii) computed at runtime.



1

2


• Gateways are decisions points used to constrain the execution flow, fork an

execution point into several or merge several into one. A gateway is represented by a

diamond and the kind of a gateway is specified by a marker.

• An exclusive gateway can be used as a decision point1 where several outgoing

sequence flows are possible. Such flows are all constrained by a condition allowing

only one of them to be used by a token. Such a condition will be evaluated based on

the process data. The gateway can be also used as a way to merge2 several sequence

flows into one. The incoming token moves straight through the gateway and goes on.


1 2


• The Exclusive Event Based gateway1 is similar to the Exclusive Data Based gateway.

The only difference is that, instead of evaluating a set of alternatives to determine

only one outgoing flow, the event based gateway will start a race between the

different events2 the process might receive, the first one to be received wins the race

and that determines which outgoing sequence flow should be used.


• Exercise: identify the event before task 2 via the BPMN poster and the specification.

1

2

2

2


• An Inclusive Gateway1 can be used as a decision point where several outgoing

sequence flows are possible, they are all constrained by conditions, each outgoing

sequence flow with a condition evaluated as being true will be followed. Effectively

it might spawn several execution points.

• Used as a merge2 the Inclusive Gateway will synchronize all the execution points

produced upstream but at most one for each incoming Sequence Flow


• A gateway's exit can also be marked with a small diagonal slash as default sequence

flow. It will be selected automatically, (only) if no condition of the other sequence

flows is true. This ensures the actual selection of at least one sequence flow.

1 2


• A Parallel Gateway1 provides a mechanism to fork and synchronize flows. There are

no conditions associated to this gateway.


1 2


• In the following model, notice how the first inclusive gateway produces two tokens

because both conditions ‘C1’ and ‘C2’ are evaluated as being true. The second

inclusive gateway2 will not only synchronize the token produced by the upstream

inclusive gateway, but also the one coming from the upstream parallel gateway

• Exercise: what happens if the second inclusive gateway is replaced by a parallel

gateway? Consider a scenario in which C1 (or C2) is false.


1 2


• The Complex Gateway was created to address complex cases which would require

the combination of several other gateways. To avoid this, the behavior of the

complex gateway can be scripted using an expression language.


• As a result the complex gateway can be used to handle every situation. Yet, best

practice is to avoid it since it makes the process models less readable.

1


• Merging alternative paths1 can also be modeled without gateways: the alternative

sequence flows directly go into the next activity2 (uncontrolled flow).

• Each single incoming token is directly processed. The activity will not wait for any

other tokens, and the tokens will not be merged.

• This behavior is appropriate for merging exclusive sequence flows where only one

token can arrive.

• Two instances of the sub-process will be created because two execution points will

reach it.

1

2

1


• Notice how the sub-process can be expanded and how the two incoming flows can

point to different start events.

http://www.iet.unipi.it/m.cimino/gpa/res/movie/mov12.swf

• Modeling with (+) or without (–) gateways?

+ it is not possible to model entirely without gateways

+ gateways are more expressive and, sometimes, necessary

– if gateways are omitted as far as possible, more compact models can be created

– gateways with several inputs and several outputs may cause misunderstanding

1

1


• Swimlanes are used to organize activities and depict collaboration between partners.

A business process is then organized inside a pool of swimlanes (business units).

• Private business processes are those internal to a specific organization, i.e. contained

in a pool, which does not interact with external pools (orchestration). A

collaboration is a synchronized interaction of more processes without a central

control (choreography), via public processes and message flows1 (white dashed

arrow) between pools (a).

• Control flow between pools (b) and message flow within a pool (c) are forbidden

1


• An example of collaboration between a patient and medical office. The processes

within pools are abstracted to hide complexity or private information to partners.

Note: pool (1), lane (2), message flow (3) and control flow (4).


1

1

2

3

4


• Lifecycle of a BPMN 2 activity/process:

• An activity is ready for execution if the required number of tokens is available to

activate the activity.

• When some data input becomes available, the activity changes from ready to

active state.

• An activity, if ready or active, can be withdrawn from being able to complete,

for tasks that are attached after an event-based exclusive gateway (race

condition): the first completed task causes all other tasks to be withdrawn.

• If an activity fails during execution, it changes from the state active to failing. If

an activity execution if interrupted, e.g. by an interrupting event or an error, its

state changes to terminating or failing, respectively.

• After all terminating requirements of a terminating activity have been fulfilled,

the state of the activity changes to terminated. Otherwise it changes to failing

and then to failed.

• If an activity execution ends without anomalies, its state changes to completing.

After all completion dependencies of a completing activity have been fulfilled,

the state of the activity changes to completed. The outgoing sequence flows

becomes active and a certain number of tokens is placed on it. Upon completion,

also data output are available.


• Only completed activities can change their state to compensating until either

compensation finishes successfully (state compensated), an exception occurs

(state failed), or termination is triggered (state terminated).

• An UML state diagram


• BPMN allows the explicit modeling of data transfer, via data objects, messages and

data store. Data objects only exist within a process, whereas data store represent

persistent data. UML class diagrams or technical terms diagrams could be used to

refine data objects in process models � process view is integrated with data view.

• A directed data association is drawn as a dotted line to model which activity outputs

and takes as input a certain data object. Data input1 and data output2 can represent a

form of dependency between activities.

• Exercise: identify the type of the first activity with BPMN poster and specification.

1 2


• Types of tasks and icons

� a service task is an automated function;

� a receive/send task receives/sends a message;

� a user task expects input by a user via a UI;

� in a business rule task some business rules are applied (e.g. via a business rule

management system launched by the process engine) to produce a result;

� a script task contains statements processed directly by the process engine;

� a manual task is carried out without IT support;

� in an abstract task no type is defined.


• Workflow patterns are models of recurrent aspects of processes control flows, as

identified by the Workflow Pattern Initiative (www.workflowpatterns.com).

• e.g., wcp-10: unstructured loop (known as arbitrary cycles): the ability to represent

cycles in a process model that have more than one entry or exit point. Block

structured looping operators (while, for,…) cannot implement this pattern.

• Example: two entry points1


1

1


• e.g., wcp-18: milestone, a task is only enabled when its process is in a specific state,

which corresponds to a specific execution point. When this execution point is

reached the nominated task can be enabled.

• If the process instance has progressed beyond this state, then the task cannot be

enabled now or at any future time (i.e. the deadline has expired).


• example: most budget airlines allow the routing of a booking to be changed

providing the ticket has not been issued.

1

1


• e.g., wcp-15: multiple instances without a priori run-time knowledge, multiple

instances of a task can be created. The required number of instances may depend on

a number of runtime factors, and is not known until the final instance has completed.

Once initiated, these instances are independent of each other and run concurrently.

At any time, whilst instances are running, it is possible for additional instances to be

initiated. It is necessary to synchronize the instances at completion before any

subsequent tasks can be triggered.


• The dispatch of an oil rig from factory to site involves numerous transport shipment

tasks. These occur concurrently, and it is always possible for additional tasks to be

initiated if there is a shortfall in transportation requirements. Once the whole oil rig

has been transported, and all transport shipment tasks are completed, the next task

(assemble rig) can commence.


• Each BPMN element has a structure made of attributes and element associations, as

defined in the BPMN specification via UML classes and XML schemas.

• Example: text annotation, used for adding explanations, with no execution effect.


• Example: timer event, which waits until (i) an absolute point of time, (ii) the end of a

certain time span, or (iii) a time cycle for a recurring time interval.


• Roughly speaking, a BPMN orchestration process (i.e., within a Pool) may be

mapped to an individual WS-BPEL process (executable on a SOA environment). A

BPMN choreography can be mapped to WSDL calls (web services interfaces).

• Not all BPMN orchestration processes can be mapped to WS-BPEL, because BPMN

allows the modeler to draw almost arbitrary graphs to model control flow, whereas in

WS-BPEL there are certain restrictions. For example, an unstructured loop cannot

directly be represented in WS-BPEL (there is no goto like statement, there are while,

for, … structured cycles)

• Further, to map a BPMN orchestration process to WS-BPEL it must be sound (i.e.,

no deadlock, no lack of synchronization, etc., van der Aalst 2003).

• A block of a BPMN diagram is a sub-diagram that is connected to the rest of the

diagram only through exactly one entering sequence flow and one leaving sequence

flow.


• Example of a data-based exclusive choice, controlled by two predicates p1 and p2,

containing three BPMN blocks G1, G2, and G31, mapped to WS-BPEL2. The

“waved rectangle” symbol is used to denote BPMN blocks.

• Example of a mapping of a simple BPMN interface, with a scalar data item

definition typed by an XML schema definition, to SW-BPEL using WSDL.

1 2


• Exercise: describe in natural language the following orchestration, concerning the

book lending process managed by a library. Use both the BPMN poster and the

BPMN specification to understand the meaning of unknown icons.


About events

• The Conditional event (a) starts a process when the related condition becomes true:

“provide additional resources when the utilization of the current ones exceeds 90%”.

• The Signal event (b) is triggered by the reception of a signal. In contrast to a message

which is always sent to a specific receiver, a signal is broadcasted everywhere, from

the same or a different pool: “when the computer system is not available, start a

process for manually processing data”.

• The Multiple event (c) combines many events; if one of them occurs, the process is

started: “a cleaning activity is started when either the cleaning interval has been

reached, or a contamination has been detected”.

• In the Parallel multiple event (d) all of the combined events must have occurred: “

the take-off procedure can be started as soon as both the permission has been given

and the engines have been checked”.


• The Terminate event (a)(c), like an un-typed end event, deletes the arriving token

when it arrives, but it also terminates the entire process, i.e., any other tokens in the

entire process are also removed at the same time.

• Example: skills and experiences of a job applicant are checked, as well as his formal

qualification is also checked in parallel. If the results of both checks are positive, the

two tokens are joined together to one token which eventually is deleted (b).

• If one of the checks has a negative result, the result of the other checks is irrelevant,

because the application will be rejected anyway. Thus the process can be

immediately terminated (a) (c).

• Replacing the terminate events by un-typed end event would cause a problem: if the

result of one of the checks would be positive and the other negative, the token from

the positive check would be stuck at the parallel gateway.


• The Signal end event (a) sends a signal, which can be received anywhere in the same

process, as well as in other processes: “once a new software version has been

released, different processes may react. Therefore, a signal is sent to everyone”.

• The Multiple end event (b) is used when the end of a process has several

consequences. When it is reached, all the combined partial events occur: “one project

has been terminated, several messages are sent”.

• In contrast to start and end events, which can only be catching and throwing,

respectively, intermediate events can either be catching or throwing. (a) conditional

intermediate event: “first the oven is turned on; the process then waits until the

temperature exceeds 180 °C, before the cake is put into the oven”. (b) signal

intermediate event: “defective sw application is shut down; then the process waits for

the signal “new version of sw available”, before the next activity can be carried out.


About triggers

• Two aspects are considered in modeling events: the trigger (cause) of the event, and

its effect in the process. Different strategies to forward the trigger to catching events:

publication, direct resolution, propagation, cancellations, and compensations.

• Typically a trigger carries information out of the scope where the throw event

occurred into the scope of the catching events. A scope is the context in which

execution happens, consisting of the set of data objects available as well as events

available for catching or throwing triggers.

• With publication a trigger may be received by catching events in any system scope.

For instance: (i) messages are triggers generated outside of their pool; (ii) signals are

triggers generated in their pool and broadcasted across processes, pools, diagrams.

• A trigger that is propagated is forwarded to the innermost enclosing scope instance

with an attached event able to catch it. For instance: (i) error triggers are critical and

suspend execution at the location of throwing; (ii) escalations are non-critical and

execution continues at the location of throwing. If no catching event is found for an

error or escalation trigger, this trigger is unresolved.

• Compensation of a successfully completed activity triggers directly its compensation

handler. Cancellation terminates all running activities and compensates all

successfully completed activities in the sub-process it is applied to. If the sub-process

is a transaction, the transaction is rolled back.


About activities

• Multi-instance activities are different from loop activities: the number of instances is

known in advance (i.e., the size of the collection of objects), and the different objects

can be processed in any sequential order (a), and also in parallel (b).

• In a process of a talent show in tv, first each candidate gives a performance; the

group of candidates is known in advance. After the last performance, the

determination of score is also sequential for each candidate. But the voting process is

carried out in parallel by each member of the jury (b).

• In an ad-hoc (sub-)process (c), marked with a tilde, internal activities are determined,

but their order is not known in advance. For instance, creative and knowledge-

intensive processes, such as computer code development (at a low level), sales

support, writing a book chapter.


• In writing a book chapter, there is no explicit process structure, but some sequence

and data dependencies can be added. Data objects as input into a task act as an

additional constraint, because the task cannot start without the appropriate input.


• Sub-processes and tasks are part of the process and cannot be used separately. A call

activity (with thick border) can be reused in processes. Call activities cannot access the

calling process, to be independent. Needed information must be provided.

• For example: the process for a treatment in hospital (in-patient) or ambulatory.


About exception handling

• While an activity / a process is carried out, sometimes an event occurs that results in

an early abortion of the activity / process.

• The sub-process “Process Order” is aborted when a cancellation is received, i.e., all

token will be removed from within that sub-process, and a token will be emitted via

the exception flow at the interrupting event.

• If the order has been completely processed, no cancellations are possible anymore. If

a cancellation arrives anyway, it will be ignored.


• Sometimes, it is desired to react to the event, but still to continue the ongoing

activity. This can be done by using non-interrupting intermediate events,

characterized by dashed borders.

• If the event “5 days after order processing has begun” is reached before the activity

has been finished, a newly created token will be passed via the exception flow to the

task “notify about intermediate status”.

• At the same time, the original token still remains in the activity “Process Order”, and

is moved via the regular sequence flow.


• Exercise: describe in natural language the following process diagram, concerning an

e-tendering.


• Consider an e-tender that sends a request for quote to multiple service providers

(e.g., warehouse storage) in a marketplace.

• The e-tender process sends out requests to each service provider and anticipates their

response through three choreography activities.

• The response branches merge at a complex gateway to model the requirement that

when 66% responses have arrived, an assessment of the tender can proceed.

• The assessment occurs after the complex gateway. If the assessment reports that the

reserve amount indicated by the customer cannot be met, a new iteration of the

tender is made.

• A key issue is to ensure that the responses should not be mixed across tender

iterations. A terminate end event ensures that all activities are terminated, when a

tender has been successful.


• The intermediate error events can only be attached to activity / process boundaries.

• In a process in a laboratory, after taking a sample, the sub-process is started. First the

sample is analyzed. This is usually followed by checking the result. If the result is

plausible, the sub process finishes. If the result is not plausible, this is an error, and

the throwing error and event “Result not plausible” is reached.

• In this case, the sub-process is aborted and the thrown error is caught by the error

intermediate event (with the same name) at the sub-process boundary. In this case,

the exception flow is followed, and the activity “Arrange for external analysis” is

triggered.


• In the sub-process, another error may occur: if during the activity "Analyze Sample"

it turns out that the equipment is defective (here also modeled with a catching error

event), it will be repaired.

• If repair has been successful, the analysis of the sample will be started again. In the

case of an unsuccessful repair, the throwing error event "Equipment out of order"

occurs.

• Again, the sub-process is aborted and the error is thrown to the intermediate error

event (with the same name) that triggers the exception flow to the activity “Arrange

for repair by Manufacturer” in addition to “Arrange for External Analysis”.

• Note that if there are several error intermediate events attached to a sub-process

boundary, they should be always labeled.

• Since unlabeled error intermediate event always catches any type of error, the

resulting implicit parallel flows are difficult to understand.

• Escalation events are rather similar to error events. Error events are mainly used for

technical problems, escalation events primarily stands for problems on a business

level, e.g., a task not completed, a goal not reached, a require agreement not

achieved.


• In contrast to errors which always abort the activity, escalations come in both

variants: the interrupting and the non-interrupting event. Non-interrupting escalation

events are regarded as the default.

• Note that if an escalation is required to abort the entire activity, the throwing

escalation event in the sub-process needs to be an end event. If a throwing escalation

intermediate event was used, the subsequent sequence flows could never be used,

because the sub-process would be aborted by the escalation.


• An Event Sub-Process is triggered by an event occurring during process execution,

and is characterized by a dotted border. The sub-process “Produce Chemicals”

contains two events sub-process, the first one is non-interrupting, i.e. the system may

be adjusted several times during production. The second one is interrupting, and also

throws an error by an end event, which is caught in turn by an error intermediate

event (Emergency shut-off) so as to emit a token to the exception flow.


About transactions

• Parts of business processes can be treated as transactions, i.e., complete units of

work. A transaction can be either carried out in its entirety, or not at all. If something

goes wrong, the transaction must be rolled back. • A BPMN transaction is drawn with a double line border. Aborting a transaction

automatically reverses the effects of the activities that have already been carried out.

For this purpose, a compensation activity is assigned to each activity that requires it.


• In the above example, the activity “Put Goods into Stock” contains a rewind-symbol

(two white arrowheads) to express it as a compensation activity. Indeed it

compensates the activity “Remove Goods from Stock”. The compensation

association is a dotted line with an arrow. Do not confuse it with message flow.

• The activity “Check order” does not have a compensating activity, because it does

not have any effects that need to be reversed.

• The throwing end event of type “cancel” is used to abort a transaction. Once the

required compensation has been carried out, the exception flow starts as the catching

cancel intermediate events.

• Note that the terminate end event is needed to cancel the token waiting for receiving

cancellations.

• Compensation can also be handled by event sub-processes. They have a

compensation start event, and can be started when a throwing compensation event of

the same name occurs after the surrounding process is already finished.

• In contrast with compensation events, error/escalation events can be caught only in

the surrounding process during its execution.


• Exercise: describe in natural language the following process diagram, concerning the

shipment process of a hardware retailer.



interaction between a pizza customer and the vendor.



incident management process.


• Exercise (reverse modeling): describe in natural language the following process

diagram, concerning the selection of a Nobel Prize in Medicine.



diagram, concerning the management of a travel reservation request by a Travel

Agency.



diagram, concerning the E-mail voting process in a workgroup.


From Reverse Modeling to Modeling

• Reverse modeling is easy: it mainly consists in translation from visual to textual

representation. In contrast, the modeling process requires a higher cognitive effort: it

is an inherently inductive (creative) process: from informal natural language to a

clear and standardized model.

• An efficient learning of the modeling process can be supported by a semi-formal

textual description. In the following, we will start to learn the modeling process via

simple scenarios described in a semi-formal textual notation.

• The semi-formal notation is made of short sentences, numbered so as to allow the

specification of the control flows with BPMN 2. It resembles detailed use cases in

UML.

• A sequential numbering is used to express a sequential order of execution. A new

level of numbering is used to express a flow internal to some branch of the control

flow. A new decimal numbering level is used to expresses exclusive branches (e.g.

x.1 and x.2). In this case, each branch starts with the if keyword. A new alphabetic

numbering level is used to express parallel or inclusive branches (e.g. x.a, and x.b).


Advertising agency

• How an advertising agency produces an advertisement, and which messages it

exchanges in doing so.

1. The Agency waits for orders from customers

2. A Customer sends an order to the Agency including the text for advertisement

3. The Agency receives the order from the Customer

4. The Agency designs the Advertisement

5.a.1. The Agency proofreads the text

5.a.2.1. If the text is okay

5.a.2.1.1. � 6

5.a.2.2. If the text is not okay

5.a.2.2.1. The Agency corrects the text

5.a.2.2.2. � 5.a

5.b.1. The Agency checks the layout

5.b.2.1. If the layout is okay

5.b.2.1.1. � 6

5.b.2.2. If the layout is not okay

5.b.2.2.1. The Agency reworks the layout

5.b.2.2.2. � 5.b

6. Wait until both text and layout are okay

7. The Agency sends a proof to the Customer

8. The Customer receives the proof from the Agency

9. The Customer sends to the Agency a feedback on the proof

10. The Agency receives the feedback on the proof from the Customer

11.1. If the Customer accepts the proof

11.1.1. The Agency archives the advertisement

11.1.2. The advertisement can be produced.

11.2. If the Customer requests changes

11.2.1. The Agency reworks the advertisement

11.2.2. � 5


Job application

• The interaction between a job applicant and an enterprise. 1. An Applicant begins to search for a job

2. The Applicant writes a job application

3. The application has been sent to the Enterprise

(The Applicant waits for the confirmation of receipt)

4. The Enterprise has received the Application from the Applicant

5. The Enterprise writes the confirmation of receipt

6. The confirmation of receipt from the Enterprise has been sent to the Applicant

7. The Applicant has received the confirmation of receipt from the Enterprise

(The Applicant waits for receiving an invitation or a rejection)

8. The Enterprise reviews the application

9.1. If the Applicant is not suitable

9.1.1. The Enterprise rejects the application

9.1.2. A message of rejection has been sent from the Enterprise to the Applicant

(The Enterprise rejected the Applicant)

9.1.3. The rejection message of the Enterprise has been received by the Applicant

9.1.4. The Applicant has been rejected by the Enterprise.

9.2. If the Applicant is interesting

9.2.1. The Enterprise Invites the Applicant to interview

9.2.2. The invitation to interview of the Enterprise has been sent to the Applicant

(The Enterprise waits for a confirmation of date)

9.2.3. The invitation to interview of the Enterprise has been received by the Applic

9.2.4. The Applicant confirm the date for the interview

9.2.5. The confirmation of date of the Applicant has been sent to the Enterprise

9.2.6. The confirmation of date of the Applicant has been received by the Enterprise

9.2.7. Time needed for starting the interview has elapsed

9.2.8. The Applicant takes part to the interview, receiving questions and providing

answers. The Enterprise carries out the interview, asking questions and

receiving answers.

9.2.9. The interview is ended.


E-mail voting system

• Management of an E-mail Voting system 1. It is Friday

2. The System receives the Issue List

3. The System reviews the Issue List

4.1 If there are no issues ready for vote

4.1.1 End

4.2 If there are issues ready for vote

4.2.1 The System announces issues for vote

4.2.2 The System sends the issue announcement to Voting Members

(The Voting Members receive the announcement from the System)

4.2.3.a.1 6 days have been elapsed since the announcement

4.2.3.a.2 The System sends to Members a deadline warning for vote

(The Members receive the deadline warning)

4.2.3.a.3 � 4.2.4

4.2.3.b.1 The Voting Members send votes to the System, the System collects votes

4.2.3.b.2 7 days have been elapsed since the start of the vote collection

4.2.4 Wait for the end of both flows 4.2.3 a and b

4.2.5 The System prepares results

4.2.6.a The posts results on Web Site � 4.2.7

4.2.6.b The System sends results of vote to Voting Members

(Voting Members receive results of vote from the System)

4.2.7 Wait for the end of both flows 4.2.6 a and b)

4.2.8.1 If enough members vote � 4.2.9

4.2.8.2 If not enough members vote

4.2.8.2.1 If members have been warned

4.2.8.2.1.1 Reduce the number of voting members and recalculate vote � 4.2.8.1

4.2.8.2.2 If members have not been warned

4.2.8.2.2.1 The System re-announce vote with warning to Voting Members

(The Voting Members receive the vote with warning from the System)

4.2.8.2.2.2 � 4.2.3.b.1


4.2.9 (go on)

4.2.9.1 If there are no issues with majority � End

4.2.9.2 If there are issues with majority

4.2.9.2.1 Reduce to two solutions

4.2.9.2.2 The System announces the vote only to Voting Members that have to change

votes (Voting Members receive the change vote message from the System)

4.2.9.2.3 � 4.2.3.b.1


From semi-formal to informal textual description

• Make a BPMN model for the following situation.

Manage orders in a manufacturing company. Order processing in a

manufacturing company takes place after an order has been received from a

customer. The order is first taken by an employee. Both manufacturing capacity and

parts availability are then checked, via appropriated business rules and an automated

information services, respectively. A rejection is sent to the customer if capacity is

not available. Otherwise, if parts are also available, an order confirmation is

automatically created and sent to the customer. Therefore, both assemblies and

invoice are then produced, by human operatives and an automated information

service, respectively. Finally, assemblies and invoice are sent to the customer. In

contrast, if there are parts to be ordered, an employee procures such parts, so as to

allow the aforementioned process to be continued, once such parts are available.

However, if some parts are still unavailable (they cannot be procured), a rejection is

then sent to the customer.

University of Pisa

MSc in Computer Engineering

Business Processes Management

“Large and complex organizations are a tangible manifestation of advanced technology, more than machinery itself.” (J.K. Galbraith)

Tutorials

Mario G. Cimino, Department of Information Engineering, Center for Logistics Systems

Pisa, March-May 2013, Monday 15.30-18.30, Room: ADInform1

Cimino – Modeling and Simulation of Business Processes using BPMN 2.0 – Tutorials – 2 of 2

Logizian 10.x

• Suite to model and simulate the execution of business processes

• Download: http://www.visual-paradigm.com/download/lz.jsp (no-install version)

• Unzip and execute bin/Logizian.exe

• Workspace: select a folder to use as workspace

• Evaluation: 30 Days Free Evaluation, Simulacian, press “Evaluate”

LogizianDrawing BPMN 2.0 business process diagram

Tutorial

http://www.visual-paradigm.com/product/lz/tutorials/businessprocessmodeling.jsp Page 1 of 11

Drawing BPMN 2.0 business process diagramWritten Date : October 30, 2009

Business Process Modeling Notation 2.0 makes a great improvement in business process modeling.There are number of new notations introduced in BPMN 2.0. In this tutorial we will show you howto draw choreography task and message. We will draw the example diagram of BPMN 2.0 draftspecification (An example of stand-alone Choreogrpahy diagram).

1. We first create a model in our project. Right click on the Model Explorer and select Model >New Model...

2. Name the model as Sales.


Tutorial


3. We can then create pools directly under Sales model without using diagram. Right click on Salesmodel and select Model Element > New Model Element.

4. You can search the model you would like to create in the New Model Element dialog. Enter Poin the Model element type text field to search for the Pool.

5. Name the pool as Customer. Press OK to proceed.


Tutorial


6. Repeat the steps to create the pool Bidder, Manufacturer and Supplier. But this time you candirectly select create pool from the popup menu.

7. Right click on the Sales model and select Diagram > Business Process Modeling > BusinessProcess Diagram.

8. We can now starts create the business process diagram. First let's create a start event. Clickon the Start Event tool in the diagram toolbar, and then click on the empty area of the businessprocess diagram.


Tutorial


9. Now click on the start event and drag out the Generic Resource icon.

10. Release the mouse button at the blank area of the diagram, and select Choreogrpahy Taskfrom the popup menu.

11. Name the task as Order Request.

12. Right click on Order Request and select Open Specification...


Tutorial


13. Specify Customer pool in Participant 1.

14. Specify Manufacturer pool in Participant 2.

15. And select Customer pool in the Initiating participant field.

16. Press OK to commit.


Tutorial


17. Now the task Order Request is done. You can see that the initiating participant is showing in thesame color of the task, and non-initiating participant is showing in gray.

18. Let's create a message between Customer and Manufacturer. Click on the Order Request taskand drag out of Association > Message icon.

19. Release the mouse button at the blank area of the diagram, and select New Message Flowfrom "Customer" to "Manufacturer".

20. Name the message as Order.


Tutorial


21. Create an Event-based Exclusive Gateway from Order Request. Click on Order Request anddrag out the Sequence Flow > Gateway icon. Drop it on the blank area of the diagram.

22. Name the gateway as Can Fulfill Order?

23. Right click on the gateway and select Type > Event-based Exclusive Decision/Merge (XOR).


Tutorial


24. Repeat the steps to complete the following business process diagram.

25. It could be multiple instance of procure parts activity doing at the same time. Let's specify it asmulti-instance activity. Right click on the Procure Parts task and select Open specification...

26. Select Multi-Instance Loop in the Loop type combo box.

27. Press OK to commit. Now the Multi-Instance loop marker is showing in the Procure Parts task.


Tutorial


28. Let's create messages between Manufacturer and Bidder on the Part Auction activity. Clickon the Part Auction activity and drag out the Message icon. Select New Message Flow from"Manufacturer" to "Bidder" from the popup menu.

29. Name the message as Part Request.

30. Repeat the step to create another message to Part Auction activity. The message should beNew Message Flow from "Bidder" to "Manufacturer". Name the message as Part Response.

31. You can see that the message from Bidder to Manufacturer is in different color since Bidder isnon-initiating participant.

32. Now the gateway All Parts Available? and All Parts Obtained? in business process diagram onlymodeled the false situation. What about the true case? It should be go to Order Confirmationactivity. We don't want the diagram become so much cross lines since they mess up the layoutand make the diagram difficult to read. We can create Link Event to simplify the flow.

33. Click on the All Parts Available? gateway and drag out the Sequence Flow > IntermediateEvent icon.

34. Name the intermediate event as A.


Tutorial


35. Right click on the intermediate event and select Trigger > Link Trigger.

36. Repeat the steps to create another link event to All Parts Obtained? gateway.

37. Click on the Link Intermediate Event in the diagram toolbar, and click on the blank area of thediagram.

38. Name the intermediate event as A.

39. Click on the link event A and drag out the Sequence Flow > Task icon, drop it to the OrderConfirmation task.


Tutorial


40. Now the business process diagram is done.

Logizian home page(http://www.visual-paradigm.com/product/lz/)

BPM tutorials(http://www.visual-paradigm.com/product/lz/tutorials/)

Visual Paradigm home page(http://www.visual-paradigm.com/)

http://www.visual-paradigm.com/product/lz/


http://www.visual-paradigm.com/product/lz/tutorials/


http://www.visual-paradigm.com/


LogizianProcess Simulation Example

Tutorial

http://www.visual-paradigm.com/product/lz/tutorials/simulacian.jsp Page 1 of 12

Process Simulation ExampleWritten Date : February 19, 2010

Simulation is a technique that helps business analysts simulates the execution of business processfor studying the resource consumption (e.g. Human resources, devices, etc.) throughout a process,identifying bottlenecks and improving the current business by process refinement and resourcesre-allocation. Simulacian is the tool that assists you to perform simulation. It includes not only thevisual simulation ability, but also charts generation functions. In this tutorial, we will study a businessprocess of performing body check, and make use of simulacian to find out the possible changes thatcan make the process more efficient.

1. Download the project Body Check.vpp attached with this tutorial.

2. Start Logizian.

3. Open Body Check.vpp

4. Open the business process diagram Body check process.

http://www.visual-paradigm.com/product/lz/provides/simulation.jsp#simulatebp


Tutorial


5. Study the process. The diagram models the current process of performing body check in ahealth center. It starts from processing a client's registration by receptionist, then the nursecheck the client's blood pressure, the X-Ray specialist test the client's lung function, and finallythe body-check officer prepare a report and present it to user.

Now, we want to know the answer of the following questions, which simulacian can help:• How long it takes to finish a process?

• What is the optimal allocation of resources like staff (e.g. specialist), rooms (e.g. Meetingroom) or devices/machines (e.g. X-Ray machine)?

• Does the process works well against a large amount of clients?

6. In order to run simulacian, we need to provide additional information to the business processdiagram, such as the current resource consumptions and durations of each activity. Right clickon the diagram background and select Diagram Type > Simulacian from the popup menu.The Simulacian Control Panel will appear for you to set simulacian details like resourceconsumption and duration of activities.


Tutorial


7. Let's start by recording the resource consumption of the current process. On the right handside of the diagram you can configure the resources that is available to the process. Availableresources mean the resources that can be used, but not necessarily be used. For example, thehealth center may have 20 wheelchairs available, while only 5 to 6 is needed. Now, click Addto add an available resource X-Ray Room, typed Room with 2 as amount. Add also a resourceMeeting Room, typed Room with 1 as amount. Other resources like wheelchairs, toilets areirrelevant to the process, and can be ignored.

8. The testing of lung function need to be done in an X-Ray room. Select task Test Lung Functionin diagram. At the Simulacian Control Panel click Add for Required Resources and select X-Ray Room from the popup menu. Keep amount be 1 as only one room is needed to perform alung function testing.


Tutorial


9. Select the task Present Report. Add Meeting Room as a required resource with 1 as amount.

10. We also need to set the duration of completing each task. Select task Process Registration.Set at the Simulacian Control Panel its duration to be 3, meaning that it takes 3 minutes tocomplete a registration process.

11. Set the duration of other tasks.

Task Duration

Check BloodPressure

2minutes

Test LungFunction

6minutes


Tutorial


Prepare Report 2minutes

Present Report 5minutes

12. The lanes such as Receptionist, Nurse, etc are the participants of the body check process. Theyare also resources that can affect the efficiency of process. We need to set the number of eachlane. Select lane Receptionist in diagram and set instance to be 1 at the Simulacian ControlPanel. This mean that one and only one receptionist is taking part in this process.

13. Set the instances of other lanes.

Lane Instance

Nurse 1

X-RaySpecialist

1

Body-checkOfficer

1

14. Click Show Inputs at the Simulacian Control Panel. Inputs are ways of running throughthe process. If a process contains gateways, you can define multiple inputs to represent theexecution of different paths due to the forking.


Tutorial


15. Take this process as example. Click Add to add an input, with name Peak hour, 10 asinstances. This means that at peak hour(s), there will be 10 clients who need to perform bodycheck. With this setting, the path (as presented in the Path list) will be executed 10 times duringprocess simulation.

16. Click the Start button to start process simulation.

17. Study the simulacian outcome. Below is a description of some of the key area that you shouldpay attention to:

No. Description


Tutorial


1 The green running circle stands for job(s) that is/are executing at the attaching task/activity.The number inside the circle stands for the number of jobs being executed.

2 The red inverted triangle stands for job(s) that is waiting to be executed. The number insidethe triangle stands for the number of pending jobs. It is attached to the end of flow to representthe need of being executed by the coming flow object.

3 The clock shows the time elapsed.

4 The Resources table show the dynamic resource consumption. It will be updated throughoutthe process simulation to reflect the use of resources at particular instant.

18. If we want to wait until the process simulation to finish, this may take about an hour since we aresimulating in real time scale. Click Stop. Change the scale to 10 mins scale and play again.

19. The pace is increased. By observing the occurrence of inverted triangle, we can determinewhere the bottleneck occur, if any, and to think of the possible improvements in further. With thisbody check process, we notice there are always pending jobs at task Test Lung Function.


Tutorial


20. Now, we need to decide how to improve the process. Since the bottleneck is at lung functiontesting, we need to decide how to speed it up. There are several possible actions we can do:• Reduce the duration of lung function testing

• Employ more specialist(s) to share the work load

Since the duration of proper lung function testing is fixed, the first action is impossible. We shalltake the second action.

21. Select lane X-Ray Specialist, change instance from 1 to 2.

22. Click Start to simulate again.

23. Strangely, the time it takes to finish the process remain the same, why? If you observe thesimulacian outcome, you can see that the task Test Lung Function is no longer the bottleneck,due to the addition of X-Ray specialist. However, the tasks Prepare Report and Present Reportnow become the bottleneck.


Tutorial


24. Let's employ one more body check officer to help. Select lane Body-check Officer, changeinstance from 1 to 2.

25. Further to employing an officer to help, we need to reserve one more meeting room. Change theamount of available meeting room from 1 to 2.

26. Simulate again.

27. It now takes 50 minutes to complete the whole process. By comparing with the original process,the improvements reduce 30 minutes operation time in total.There are other possible solutions instead of employing more staff. For example, we cancomputerize the process by not presenting report to client, but to print it out with cleardescription on each checked item, and let the client read himself. We may also buy advanced X-Ray machines to speed up the lung function testing process. Simulacian does not recommenda solution to you, but to let you execute your idea of improvements easily through updatingdiagram content and resource allocation, and give you figures to show the effect of changes, tohelp you decide the best solution.


Tutorial


28. Sometimes, just by watching the simulacian outcome is not enough in finding out the bottleneck,especially when the diagram is large, and have many, many bottlenecks. In such cases, you canproduce charts for simulacian outcome, which helps quantify resource consumption and queuingtime for each activity. Click Simulacian Charts in Simulacian Control Panel.

29. Study the chart in Completion page. The chart show the status of inputs completion againsttime. Remember we have set 10 instances to input Peak hour? From the chart you can see thecompletion of the 10 instances throughout the process simulation. As time goes by, there will bemore inputs being completed, and less inputs being processing.


Tutorial


30. Open the Resource Usage page. The chart here shows the level of resource consumption.In this example, both X-Ray room and meeting room are fully used (100%) within the process,which is ideal. But if you try to add 3 to 4 available X-Ray rooms, and review the chart, you willsee that the peak usage will be less than 100% because some rooms are idles throughout thewhole process. By studying this chart, you can adjust the amount of resources to apply to theprocess.


Tutorial


31. Open the Queue Time page. The chart here shows the time the flow objects spent on waiting,which corresponds to the time an inverted triangle appear during simulacian. By studying thechart, we know that process registration now takes over 10 minutes in waiting, prepare andpresent report still take over 4 minutes after the improvement. By studying this chart, you candecide whether further improvements need to be carried out.










LogizianDevelop as-is and to-be business process

Tutorial

http://www.visual-paradigm.com/product/lz/tutorials/developasistobebpm.jsp Page 1 of 8

Develop as-is and to-be business processWritten Date : January 31, 2012

A business is almost always looking to maximize profitability. In the broadest sense, a businessusually aims to achieve that by trying their best to minimize cost or maximize revenue, or probablyboth. So companies are highly encouraged to improve how business runs in hopes to improveproductivity or deliver services more efficiently.

Business process improvement may come in some of the common forms below:

1. Technology adoption. For example, old-style stores perform operations by human. If wecomputerize a business for higher efficiency, it is very likely that the process will need to be re-engineered in order to reap the benefit from adding the computer system.

2. Changes in supplier/manufacturer/customer relations which require business process re-engineering. Perhaps customers are dissatisfied with the slow courier service. So you mustchange the courier company to avoid losing customers. Another case might be that the previoussupplier which used to deliver goods to you weekly cannot accommodate your new dailyschedule and leaves you no choice but to seek a new supplier that does. And it would probablycall for a process change in order to work with the new supplier.

3. Change of management direction and process obsolescence. Perhaps the management wantsto close down a department. If a process relies on that department, you have no choice but tomodify the process.

Attempting to improve a process begins with understanding how it works currently. Why not describethe process visually using graphical notation of a widely adopted industry standard? A visualrepresentation with a common ground means less confusion and saves more time for the actual re-engineering part. And that standard is called Business Process Modeling Notation (BPMN). In BPMN,the current process to be enhanced is called as-is process. After it gets re-designed, we call it to-beprocess.

Software tool we'll be usingIn this tutorial, we will be using Logizian to create BPMN-compliant business process diagrams. Itprovides a robust means to capture the current business process, design enhancement and identifythe enhancement between the current and enhanced processes.

Capture the Current Process in As-is Process

http://www.bpmn.org/



Tutorial


1. In the Model Explorer, right click on untitled and select Model > As-is Process.

2. In the Model Explorer, right click As-is Process and select Diagram > Business ProcessModeling > Business Process Diagram.


Tutorial


3. The example through which we'll demonstrate is about an online shop selling goods. Theprocess begins with the sales representative receiving a purchase order from a customer andproceeds to check stock level. If there is enough stock on hand to meet the order, the salesrepresentative will pack them. The process ends with shipping them along with an invoice.In case of insufficient stock, the sales representative will suggest the customer to amend thepurchase order.

We are not going to walk you through how to draw business process diagrams using BusinessProcess Modeling Notation (BPMN) here as we have another tutorial that covers it. We alsohave a tutorial video called Introduction to BPMN - Swimlanes that you might want to have alook.

Re-design to Enhance the Current Process in To-be ProcessLet's just say that our business has grown so much that we now have a warehouse to keep ourstocks. So we are looking to improve our current (as-is) process to best use the new resource. Andwe will show an example of modeling the enhancement below in a to-be process diagram.

http://www.visual-paradigm.com/product/lz/tutorials/bpmn1.jsp

http://www.youtube.com/watch?v=m7c--hHlZBU


Tutorial


1. First, create a to-be process diagram from the current process diagram. Right click any whitespace in the background of the as-is process diagram and select Utilities > Create To-beProcess.

You'll see that a new diagram is created containing the to-be process.Note: Initially, all model elements are selected in the new diagram. To deselect, click any whitespace in the background.


Tutorial


2. Add a lane to Online Shop. Name it Warehouse. In the Sales Representative lane, selectPack Goods, Ship Goods and Order completed and drag them to the Warehouse lane

3. Next, we are going to insert a task between Sufficient Stock? and Pack Goods. Place yourmouse pointer over the Yes flow coming from Sufficient Stock?. Click on the bubble to showthe little cube icon. Click it to show the popup menu and select Task. Name the new taskRequest Pack Goods.

The completed to-be process diagram should look something like this.

Compare Changes in As-is Process and To-be ProcessTo compare differences between two diagrams visually, we can make use of a function called VisualDiff....


Tutorial


1. From the main menu, select Modeling > Visual Diff....

2. In the Visual Diff dialog box opened, select Transitor for Strategy. For Compare, select ModelElement to be the level at which we compare our diagrams.

3. Check the checkboxes next to Use Working Project. Select As-is Process on the left. SelectTo-be Process on the right.


Tutorial


4. Click Compare at the lower right corner to see differences being "spotted."For example, from the list below the diagrams, we know that changes were made toPack Goods as it says Pack Goods (Modified). If you click on the row, you will see thecorresponding changes in purple in the diagrams.

New tasks will be noted as well. Remember (in the enhancement) we added a new task for thesales representative to request the warehouse to pack goods? That is also "spotted" in our list.


Tutorial


5. If you want, you can keep the comparison results in a PDF file for future reference. Simplyclick Export PDF... at the lower left corner and give the PDF file a name (e.g. comparisonresults.pdf).

Download: sample project file & PDF file for comparison results

Related Links1. Tutorial video - Introduction to BPMN Written tutorial - Introduction to BPMN Tutorial video - how

to simulate business processes to quantify performance




http://s1.linkvp.com/bpva/tutorials/developasistobebpm_screenshots/20120131/resources/Online-Shopping_tobe.vpp

http://www.visual-paradigm.com/?format=pdf

http://www.youtube.com/watch?v=m7c--hHlZBU

http://www.visual-paradigm.com/product/lz/tutorials/bpmn1.jsp

http://www.youtube.com/watch?v=b_47NV2ibzw

http://www.youtube.com/watch?v=b_47NV2ibzw







Search...

Home Products LZ Tutorials Simulate Processes to Quantify Performance

What Logizian Provides What's New Tutorials Editions Resources

Relatively speaking, while it may take "little" money to design a business process, it takes a whole

lot more for implementation. So it is immensely important that we think twice to make sure that it is

worthwhile to do so beforehand. This is where computer simulation comes in to help via load and

performance testing. They give us some ideas as to how the system or process would behave

under the scenarios we defined, quantitatively, in terms of cost and time investment.

Simulate Processes to Quantify Performance

In this tutorial, we will show you how to set parameter values and create scenarios in order to run simulation. During the simulation, you will

see that the values on the time and cost meters change as the process moves along under those scenarios. At the end, you will find out the

total time and cost the process would require for completion, which give us a means to quantify the performance. If we simulate another

process design for the same goal and scenarios, we would have some metrics to compare the robustness among our different process

designs.

Modeling Software Tool

Logizian Simulacian Edition from Visual Paradigm

A Sample Business Process

As an example, we will look at the process of applying for a credit card at a fictional bank called ABC Bank. It involves the applicant, the front

desk and back office of the bank. The process begins with the front desk helping the applicant filling an application form. Completed

applications then go to the back office for verification. If the application is found to be lacking some supporting documents, the back office will

request the applicant for them. When there is sufficient information, the application will be processed. If it gets approved, a credit card will be

issued. If not, a rejection letter will be sent to the applicant for notification.

Download sample project file: Samples_Simulacian.vpp

Note: We are not going to show you how to draw the business process diagram step by step here because we have another tutorial which

covers that. If you would like to learn more about the drawing part (with BPMN), please read our tutorial of Introduction to BPMN Part I.

User Rating: / 3

January 26, 2012

Views: 2,636

PDF Link Add comments

Simulate Business Processes to Quantify Performance http://www.visual-paradigm.com/product/lz/tutorials/simulateprocesses.jsp

1 of 9 11/03/13 14:43

Steps to Set Up Process Simulation

Open the sample project file in Logizian Simulacian Edition.

Along with the business process diagram, we will be working with the simulation pane below.

If the simulation pane is not open already, from the main menu, select Modeling > Simulacian...

Set Time and Cost for Tasks

For each task, we will set the time and cost required to complete it once. For example, select the task called Assist application to fill

application form. For duration, enter 30 in the box to the left of the letter m (for 30 minutes). For cost, enter 10.0 for ten U.S. dollars.

The default currency is USD. But you can always change it to tailor your need.

1.

Share Select Language ▼ BP-VA 10.0 Build sp1_20130102 ChangeLog


2 of 9 11/03/13 14:43

For the remaining tasks, you could use the sample values below or enter your own.

Assist applicant to fill application form (30 minutes, USD10.0)

Verify application form and supporting documents (10 minutes, USD3.0)

Request more supporting documents (5minutes, USD3.0)

Process application and suggest credit limit (1 hour, USD30.0)

Send rejection letter (10 minutes, USD5.0)

Issue credit card (1 hour, USD10.0)

Close application (zero time, zero cost)

Set Lane Instance

In this section,we will define resources for both the front desk and the back office. First, select the lane called Front desk and enter 1 for

Instance (e.g. manpower).

Then select the lane called Back office in the diagram and enter 1 for Instance (e.g. manpower).

2.



3 of 9 11/03/13 14:43

Create Scenarios

In this section, we are going to create 2 scenarios. First, click any white space in the background of the business process diagram. That

should open the Scenario pane.

Click Add to create a scenario which appears in row in the table. Basically, three things define a scenario - (1) its name, (2) the number of

cases and (3) the path.

Name the new scenario enough info initially and approved. Say there are 30 credit card applications in this scenario. So enter 30 for the

number of cases.

Set the following path to account for the scenario in which there is sufficient information when the application is first received and that it

gets approved in the end.

3.



4 of 9 11/03/13 14:43

Tip: If you check the checkbox next to highlight in diagram, you will see the path visually in the business process diagram in highlight

(e.g. purple) as you define it in the list.

Then create another scenario called requested more info and approved to account for applications that didn't come in with sufficient

information initially and only get processed after requesting for additional information. But in the end, they get approved.



5 of 9 11/03/13 14:43

Simulate

After setting up the scenarios, we are now ready to simulate them. To start it, simply click the start button.

You are free to pause the simulation any time by clicking the pause button. For example, so far the process has taken 0 days 7 hours 30

minutes 0 seconds. And a cost of USD320 has been accrued.

During the simulation, you will see 2 symbols in the diagram - red inverted triangles and green gears. Take the below diagram as an

example. There are 2 applications (after requesting and receiving additional supporting documents) now pending to be processed for a

credit limit. Meanwhile, at this very moment, there are 6 applications (that came with enough information when first received) now pending

to be processed, just like the other 2 waiting in queue. The green gear shows that there is one application currently being processed by that

task.

As the simulation ends, you can read the meter and see the total time and cost for that process (e.g. 3 days 4 hours 45 minutes 0 seconds

and USD1,870). The red inverted triangles and green gears should all have disappeared by now.

4.



6 of 9 11/03/13 14:43

Show Charts

The simulation charts tell us statistics about things such as resource consumption, which helps us identify bottlenecks and improve the

process.

Click the Simulacian Charts button.

Various charts appear in the window opened. For example, we can choose to look at the Cost Per Flow Object chart and learn which task

is most expensive in terms of cost (e.g. Process application and suggest credit limit).

Let's have a look at another chart called Queue Time.

The chart shows the total queue time of applications waiting in line to be processed by various tasks within the process.

5.



7 of 9 11/03/13 14:43

It turns out that pending applications were in line the longest for the task to issue a credit card (e.g. a bottleneck).

To improve our process, we might consider redistributing the task load. The top 3 tasks in terms of queue time would be a logical place to

begin playing around with. Unfortunately, the top two tasks - issue credit card and process application and suggest credit limit - must

stay in the back office due to sensitive information. So the next best try would be to move the task to verify application form and

supporting documents to the front desk.

After we move the task to the front desk lane and rerun the simulation, we learn that the new process would save us a little over 10 hours

of work only, which may not worth the trouble to deploy a process change.



8 of 9 11/03/13 14:43

Cimino – Modeling and Simulation of Business Processes using BPMN 2.0 – Tutorial 5 – 1 of 15

The process of bag manufacturing

• The figure outlines the macro processes of bag manufacturing in a workshop. First,

cutting and preparing components, where semi-finished products originate; then,

assembling and checking against quality. If products are good, they are packed and

shipped out. Otherwise corrective actions are triggered to handle error (not modeled).

• In the two exclusive gateways, the make-or-buy business decision is made, by

comparing the costs and benefits of carrying out internal or external manufacturing

of product components, via outsourcing to a third party specialist.


1. Select each task, and insert duration and cost per execution.


2. Data concerning all the activities of the model.

Activity Average Duration Average Cost (EUR)

Cutting 25 16

Preparing Components (internal) 28 31

Preparing Components (external) 24 48

Assembling (internal) 93 67

Assembling (external) 68 93

Quality check and packing 42 26

Outsourcing Management * 5 2

* Interfacing with the third party

3. Define number of available instances (pools)1, time scale 2 and currency3.


4. Create a scenario: left-click on the diagram background and press add;

5. Insert the name (Internal production) and the number of cases;

6. Create the path of the scenario.


7. Two scenarios: internal production and (partially) external production.


8. Click the Play/Stop/Pause buttons to simulate;

9. Look at the red inverted triangles (tokens queues) and at the the green gears

(processing tokens);

10. Look at the final duration and cost;


11. The two scenarios are executed considering the related number of tokens, e.g., 30%

and 70%;

12. Click on the logo for plotting important duration and cost parameters;

13. Completion against time: to be processed, processing, and processed tokens:


14. Resource usage against time:


15. Queue time


16. Cost per flow object (cost per use)


17. Other costs

18. Time Cost


19. Exercise

Suppose to aim at producing 40 bags, by combining internal and external production

scenarios, with a single bag workshop and a single third party specialist.

Simulate:

a) a scenario of totally internal production; analyze results/plots;

b) a scenario of partially external production; analyze results/plots;

c) find a combination of the two scenarios so as to carry out the process in both less

time and cost, with respect to (a) and (b).


Solution

Let x1 be the percentage of bags that are produced internally, and y(x) the total

duration of the production process of 40 bags.

a) x1=100% → 5g 5h 20' 5600€, with maximum queues on the last phases, due to

the sequential character of the workflow;

b) x1=0% → 3g 6h 52' 7640€, with maximum queues on the assembling stage,

which is a process with larger duration, with respect to the other processes;

c) by carrying out internally some units, the assembling is parallelized, thus reducing

queuing effects, and then saving total time and cost; e.g.. x1=20% → 3g 2h 38'

7232€;

d) In the context of luxury handbags production, for a given quality level that is

guaranteed by the control quality process, the total duration of the process is the

main Key Performance Indicator (KPI), rather than the total production cost;

e) Is 20% the best solution in terms of total duration? Assuming that there is a

unique minimum, it can be efficiently found by using a binary search;

f) Given x1 and x3, calculate the total duration of the process for x1, x3, and for the

center x2=(x1+x3)/2, as well as for a value very close (dx) to each of these points.

On the basis of the descent direction we can establish the position of the

optimum with respect to the center.


g) e.g. if the total duration goes down on the right side of x1, and on the left sides of

x2 and x3, then the minimum is between x1 and x2.

h) By carrying out 8 simulations, it can be determined that the optimum is located at

x=12-13%, i.e., 5 bags produced internally and 35 externally, with a total

duration of 3d 16m, and a total cost of 7385€.


Selecting the best configuration for a Hospital Emergency Center

a. Consider the operation of a Hospital Emergency Center (HEC). The process begins

when a patient arrives through the Acceptation process in the Entrance Room of the

HEC, and ends when a patient is either released from the HEC or admitted into the

hospital for further treatment.

b. Patients arriving on their own, after acceptation, sign in, and then are assessed in

terms of their condition (triage). Depending on their condition, patients must then go

through the registration process and through the treatment process before being

released or admitted into the hospital.

c. Arriving patients are classified into different codes (levels), according to their

condition. With Red Code, patients are more critical than with Yellow and Green

codes. Indeed, Red Code patients are taken to an Emergency Room (ER)

immediately after acceptation. Once in the room, they undergo their treatment.

Finally, they complete the registration process before being either released or

admitted into the hospital for further treatment. Patients arriving by ambulance are

directly classified with Red Code.

d. Yellow and Green code patients must first sign in with an Administrative Clerk.

After signing in, their condition is assessed by a Triage Nurse, and then they are


taken to an Emergency room. Once in the room, Yellow and Green code patients

must first complete their registration, then go on to receive their treatment, and,

finally, they are either released or admitted into the hospital for further treatment.

e. The treatment process consists of the following activities:

1. A secondary assessment performed by a nurse and a physician.

2. Laboratory tests, performed by a patient care technician.

3. The treatment itself, performed by a nurse and a physician.

f. The registration process consists of the following activities:

1. A data collection activity performed by an Administrative Clerk.

2. An additional data collection activity performed by an Administrative Clerk, in

case the patient has Worker’s Compensation Insurance.

3. A printing of the patient’s medical chart for future reference, performed by an

Administrative Clerk.

g. Finally, 90% of all patients (regardless of the assigned code) are released from the

HEC, while the remaining 10% are admitted into the hospital for further treatment.

The final release/hospital admission process consists of the following activities:

1. In case of release, an administrative clerk fills out the release papers.

2. In case of admission into the hospital, an Administrative Clerk fills out the

patient’s admission papers. The patient must then wait for a hospital bed to


become available (Admission). Finally, the patient is transferred to the hospital

bed.

h. The HEC has the following resources: nurses, physicians, technicians,

administrative clerks, Medical and Administrative rooms.

i. Due to cost and layout considerations, hospital administrators have determined that

the staffing level must not exceed 7+2 nurses, 3+1 physicians, 4 technicians, and

4+4 Administrative Clerks.

j. Furthermore, the HEC has 7 medical and 13 administrative rooms available;

however, using fewer rooms would be beneficial, since other departments in the

hospital could use the additional space more profitably. The hospital wants to find

the configuration of the above resources that minimizes the total asset cost. The asset

cost includes the staff’s hourly wages and the fixed cost of each room used. We must

also make sure that the total cost is lower than 300K$ and that, on average, patients

do not spend more than 6 hours in the HEC.

k. Summary of parameters and constraints

Resource or Activity Data or Constraints

Admission 90% of all patients are released

10% are admitted for further treatment


By Ambulance 5.6 % of all patients

RED code 14.8 % of all patients

Available Nurses 7+2

Available Physicians 3+1

Available Technicians 4

Available Administrative Clerks 4+4

Available rooms 20 (using fewer rooms would be beneficial)

Staff’s hourly wages ($):

Nurse: 54.10

Physician: 78.70

Technician: 36.54

Admin: 29.03

Fixed Cost of each room used ($) 55800 (Medical Room)

13200 (Administrative Room)

Max Average Time spent per patient Less than or equal to 2h 24 m

Acceptation 30s / 0.24$ (1 Admin.)

Sign in (average duration/cost) 3m 12s / 1.55$ (1 Admin.)

Triage (average duration/cost) 5m 24s / 11.95$ (1 Nurse + 1 Phys.)

Registration (average duration/cost) 9m 30s / 4.59$ (1 Admin.)

Admission (average duration/cost) 3m 42s / 1.79$ (1 Admin.)

Release (average duration/cost) 2m 12s / 1.06$ (1 Admin.)

Treatment (average duration/cost) 25m 54s / 80.69$ (2 Nurses + 1 Phys.)

Arrangement (average duration/cost) 21m 15s / 10.28$ (1 Admin.)


l. An example of BPMN process layout:


Exercise:

a. Create all scenarios according to the above constraints, for 100 patients.

b. Try the following configuration, and compute the average patient cycle time.

N. of Lanes Related Activities and required resources

1 Entrance Acceptation, Arrangement: 1 Admin Clerk + 1 Admin Room

3 Administrative

Service

Sign in, Registration, Admission, Release:

1 Adm Clerk + 1 Adm Room

1 Triage Triage: 1 Nurse + 1 Physician + 1 Medical Room

2 Emergency Treatment: 2 Nurses + 1 Physician + 1 Techn + 1 Med Room

c. Discover other configurations, able to decrease the average patient cycle time

d. Redesign some features of the model to improve the cycle time of patients. In the

new model, we assume that patients can go through the treatment process and the

registration process in parallel. That is, we assume that, while the patient is

undergoing treatment, the registration process is being done by a surrogate or

whoever is accompanying the patient. If the patient’s condition is very critical,

someone else can provide the registration data; however, if the patient’s condition

allows it, then the patient can provide the registration data during treatment.

e. Compute the Total Asset Cost for this configuration, and discover some optimal

solution that minimizes the average patient cycle time.


Selecting the best configuration for a Hospital Emergency Center - Solution

a. The first step it to determine the number of tokens for each scenario. This can be

done with simple arithmetic, as represented in next pages.

b. The following resources have been assigned to the HEC:

N. of Lanes Related Activities and required resources

1 Entrance Acceptation, Arrangement: 1 Admin Clerk + 1 Admin Room

3 Administrative Service

Sign in, Registration, Admission, Release:

1 Adm Clerk + 1 Adm Room

1 Triage Triage: 1 Nurse + 1 Physician + 1 Medical Room

2 Emergency Treatment: 2 Nurses + 1 Physician + 1 Techn + 1 Med Room


• Total Duration and Total Variable Cost: 21h 40m 12s 10037$

• Looking at the Completion chart, it can be assumed that the completion time for each

scenario is linearly increasing. Thus, the following completion data can be derived:

Scenario First Token

exits at

Last Token

exits at

N. of

Tokens

Average

Completion

time

OWN & NORED & REL 2h 40m 21h 40m 77 12h 10m

OWN & RED & REL 3h 25m 5h 5m 8 4h 15m

OWN & NORED & ADM 3h 35m 7h 25m 8 5h 30m

OWN & RED & ADM 4h 4h 1 4h

AMB & RED & ADM 2h 35m 2h 35m 1 2h 35m

AMB & RED & REL 2h 20m 3h 15m 5 2h 47m

AVERAGE TIME

• Using Microsoft Excel, the total average time is 10.35 hours, very higher than 2.4!


• The maximum resource usage can be also derived. Looking at the Resource Usage

diagram (expressed in percentage w.r.t. the available ones) the maximum usage per

resource can be easily calculated:

Nurses: 5/7; Phys: 3/3; Techn: 2/4; Adm: 4/4; Med Rooms: 3/7; Adm Rooms: 4/13;

• Total costs = Fixed costs + Costs per Input =55800*3 + 13200*4 + 10037= 230237$


c. Looking at the Queue Time, it can be seen that the bottleneck is at the Treatment

activity. Let us increase the number of resource for treatment: 3 Emergency Room.

• Total Duration and Total Variable Cost: 17h 21m 12s 10037$

• Total Average Time: 8.65h

• The new Queuing situation is more balanced.


Simulation and Emulation

• Simulate the following simple diagram, with a unique scenario of 10 instances, so as

to understand how the simulator works with parallel flows.

• Create a unique scenario by using parallel gateways only (fig.)


• As expected, the parallel gateway duplicates the number of tokens. Indeed, the

maximum number of tokens produced during the simulation is 20.

• To allow an actual parallel execution of Task2 and Task3, the number of instances of

the Pool has to be increased, e.g. to 4. The total duration is 1h 40m.

• Note that the simulator engine may perform an unbalanced allocation of Pool

resources to tasks when the number of tokens is low (a few units). For example, in

some of such trials we may discover that all Pools have been moved from Task2 to

Task3 (and vice versa), leading to an increase of the completion time.

• To avoid these occurrences, perform the same trial varying the number of resources.


• For more complex layout, simulation of parallel paths in the same scenario is not

allowed by the simulator. This means that, when creating a scenario, you are asked to

choose a path even using the parallel gateway.


• For this reason, you have to create an emulation diagram, i.e., a diagram which is

purposely created by considering the limitation of a simulation engine, so as to

follow the actual model.

• Simulation involves modeling the underlying state of the system. Ideally, you

should be able to look into the simulation and observe properties that you would also

see if you looked into the original system.

• In practice, there may some shortcuts to the simulation for performance reasons, that

is, some internal aspects of the simulation may actually be emulated.

• Emulation is the process of mimicking the outwardly observable behavior to match

an existing system. The internal state of the emulation mechanism does not have to

accurately reflect the internal state of the system which it is emulating.

• Emulate the above example by using inclusive/exclusive gateways and by combining

two scenarios.


• With the emulation, the total duration is shorter, i.e., 1h 20m, because there are no

synchronization constraints between Task2 and Task3 at the merging node.


• When assigning multiple instances of a resource (lane), the execution of single tasks

(e.g., Task1 and Task4) is also parallelized, because the simulator allocates the

available resource to all tasks.

• In order to decide which tasks have to be parallelized, we can define two levels of

parallelism with a mirror resource, equipped with more instances, as the resource

for parallel segments. A separate base resource, with a single instance, devoted to

single task execution.


Emulating interarrival time

• When simulating no-peak scenarios, tokens (e.g. customers) arrive at time intervals.

The average time between two consecutive customer is called interarrival time, and

can be emulate with a 1-instance lane with a timer event.

• Example1: with T1 the total duration is 1:20:00

• Example2: with T2 the total duration is 1:41:00

• However, the interarrival time should be subtracted from the total duration.


Example 1


Example 2


Example 1


Example 2


Example 1


Example 2


Selecting the best configuration for a Hospital Emergenvy Center – Part d-e

• An example of redesign.

• The number of tokens for each scenario can be calculated with the methodology

already employed in the first model.


• This model cannot be simulated, because the tool does not allow the definition of

scenarios with multiple paths at the parallel gateway, and with the inclusive merge

node. An emulation model must be performed.


• An emulation model. The model is redundant, because there are two occurrences of

Registration and Treatment tasks. The activity Treatment-with-Registration is an

emulation of the parallelism between the two activities.


• Treatment-with-Registration:

Duration = the maximum of the durations (of Treatment and Registration)

Resource = the union of the resources

Cost = the sum of the costs

• Total duration and cost in the emulation model: 17h 31m 54s 10037$.

• The total duration is higher than the duration of the simulation model (17h 21m 12s).

Explain the reason.


Solution

• The emulation model requires more human resources to be experimented, because it

exploits parallelism in the Emergency Room.

• Compare both models as follows:

MODEL RESOURCES RESULTS COMMENTS

Simulation 4 Physicians

4 Emerg. Rooms

14h 45m 18s

10037$

The maximum number of parallel

execution in Emergency Room is

still 3. Why? Because 8 Nurses are

needed, instead of 7.

Simulation 4 Physicians

4 Emerg. Rooms

8 Nurses

13h 04m 12s

10037$

Simulation

(best)

20 Units of each

human resource

11h 23m 36s

10037$

Tot avg Time

6,18

Tot Costs

55800*5+

13200*4+

10037 =

341837$

Theoretical experiment, with infinite

human resources.

It is used to assess the resources

actually needed: 9 Nurses, 5 Phys, 4

Tech, 4 Adm, 5 Med Room, 4 Adm

Room.

Start-End Completion time per

scenario: (1) 2:20-11:25; (2) 3:20-

4:30; (3) 3:08-6:10; (4) 4:00-4:00;

(5) 2:45-2:45; (6) 2:26-3:20.


MODEL RESOURCES RESULTS COMMENTS

Emulation 4 Physicians

4 Emerg. Rooms

14h 56m 30s

10037$

The maximum number of parallel

execution in Emergency Room is

still 3. Why? Because 8 Nurses are

needed, instead of 7.

Emulation 4 Physicians

4 Emerg. Rooms

8 Nurses

13h 17m 48s

10037$

Emulation

(best)

20 Units of each

human resource

11h 11m 12s

10037$

Tot avg Time

5,75

Tot Costs

55800*4+

13200*4+

10037 =

286037$

Theoretical experiment, with infinite

human resources.

It is used to assess the resources

actually needed: 9 Nurses, 5 Phys, 4

Tech, 8 Adm, 4 Med Room, 4 Adm

Room.

Start-End Completion time per

scenario: (1) 2:09-11:12; (2) 1:49-

3:32; (3) 2:34-5:02; (4) 2:12-2:12;

(5) 1:18-1:18; (6) 0:58-1:40.


Modeling from informal diagrams: selecting the best staffing level for a claims

process at an insurance company

• A personal claims department in an insurance company handles claims made by their

clients. Figure on the next page, is a non-BPMN process map depicting the personal

claims process in terms of swimlanes.

• The first lane corresponds to work done by a claims handler (CH) located at the

client’s local service center. Upon arrival of a claim, the assessor determines if the

client has a valid policy. If no (5% of all cases), then the case is terminated;

otherwise (95% of all cases), the assessor enters the appropriate information in the

system.

• In the second lane, an assessor located at the service center (ASC) receives data from

the claims handler. The assessor first determines if the claim is covered by the

client’s policy. If not (5% of cases), the case is terminated; otherwise (95% of cases),

the assessor approves the preliminary estimate of the damage. If the damage exceeds

$2,000 (35% of cases), the claim is sent to an assessor at headquarters for approval;

otherwise (65% of cases), it is sent directly to a Senior Assessor.


• Lane 3 corresponds to the assessor at headquarters (AHQ). The assessor first

authorizes the on-site investigation of the accident. If the investigation determines

that the incident is not covered by the client’s policy (2% of cases), then the case is

terminated; otherwise (98% of cases), a final price is determined and the case is

approved.

• In lane 4, the senior assessor (SA) receives the claim, checks it, completes it, and

provides the final approval. Once the claim is approved, it is sent to documentation

control.

• Documentation control (DC), in lane 5, is in charge of processing the payment to the

client, closing the case, and, finally, filing the claim.

• The problem in this example is to find the most efficient staffing levels for each of

the five resource types. Each resource type has a maximum limit of 20 people, and

the overall headcount in the process cannot exceed 70. For 100 claims in a peak

scenario, the total duration should be lower than 200 minutes

• Perform a what-if analysis of possible solutions to this problem. To decide which

configuration best aligns with service levels and process goals, analyze the trade-offs

between headcount and total duration. Finally, consider also sensitivity: which type

of resource produces a lower difference in total duration when reduced by a unit.


• Scenarios

SCENARIO N. OF CASES

1) POL-NOT-VAL 5

2) POL-VAL & INC-NOT-COV 5

3) POL-VAL & INC-COV & >2000 & INC-NOT-COV 1

4) POL-VAL & INC-COV & >2000 & INC-COV 31

5) POL-VAL & INC-COV & <2000 58

TOTAL CASES 100

• Duration

ACTIVITY AVG. DURATION

RECEIVE CLAIM 2m 12s

ENTER DATA INTO SYSTEM 10m 32s

INVESTIGATE 19m 28s

APPROVE PRELIMINARY ESTIMATE 3m 54s

AUTHORIZE ON-SITE INVESTIGATION 2m 46s

DETERMINE PRICE OF DAMAGE 37m 16s

APPROVE CASE 2m 10s

CHECK CLAIMS CASE 3m 38s

APPROVE CLAIM 1m 18s

SEND PAYMENT TO CLIENT 7m 11s

CLOSE CASE 1m 46s

FILE CLAIM 3m 22s


• Optimization problem: minimize the headcount, i.e., the total number of lane

instances, under the following constraints: (i) each lane type has a maximum limit of

20; (ii) the total number of lane instances in the process cannot exceed 70; (iii) the

total duration should be lower than 200 minutes.

• An experiment with 20 instances available for all lanes (20, 20, 20, 20, 20) produces

a total duration of 3h 26m 5s (206 minutes).

• The longest queues are located in the second lane (Fig.1, investigate, approve

preliminary estimate), i.e., “Assessor Service Ctr”. Moreover, the second lane

exploits all 20 instances for a considerable amount of time (Fig.2, green line). Thus,

a reduction of the available instances of the second lane may produce a substantial

increase in the total duration. Second lane => 20 instances.

• Queues in the second lane may also depend on the inter-arrival time, which is

determined by resources in the first lane. There could be a positive effect in the

reduction of some resource in the first lane:

Inst. 19 18,17,15 16 14 13,12 11 10 9 8

Dur. 3:18:55 3:11:05

(191m)

3:12:13 3:13:17 3:15:29 3:14:55 3:12:05 3:11:59

(192m)

3:24:01

• First lane => 9 instances (9, 20, 20, 20, 20). New results are shown in Fig. 4, 5, 6.


Fig. 1 – Queue Time with (20, 20, 20, 20, 20) instances


Fig. 2 – Resource Usage with (20, 20, 20, 20, 20) instances


Fig. 3 – Completion Time with (20, 20, 20, 20, 20) instances

Note that scenarios 4 and 5 (the most relevant in terms of tokens) cannot be carried out in

parallel: when there is an increase in the completed instances of a scenario, the other

scenario is stationary, and vice versa.


Fig. 4 – Queue Time with (9, 20, 20, 20, 20) instances


Fig. 5 – Resource usage with (9, 20, 20, 20, 20) instances


Fig. 6 – Completion Time with (9, 20, 20, 20, 20) instances

Note that scenarios 4 and 5 (the most relevant in terms of tokens) can be carried out in

parallel: both scenarios can increase the completed instances.


• The activities included in the IV lane (Senior Assessor) do not produce significant

queues and the maximum resource usage is 55-60% (Fig.5). Thus, we can reduce the

available instances without effects on the total duration:

Instances 19,…,6 5

Duration 3:11:59 3:15:53

• Fourth lane => 6 instances (9, 20, 20, 6, 20).

• Similarly, the maximum resource usage of the III lane (Assessor Headqtr) is about

90% (Fig.5). Thus, we can reduce the available instances without effects on the total

duration:

Instances 19,18 17

Duration 3:11:59 3:13:05

Third lane => 18 instances (9, 20, 18, 6, 20).

• Again, the maximum resource usage of the V lane (Document Control) is about 85%

(Fig.5). Thus, we can reduce the available instances without effects on the total

duration:

Instances 19,..16 15

Duration 3:11:59 3:12:36

• Fifth lane => 16 instances (9, 20, 18, 6, 16). Total duration 3:11:59


Fig. 7 – Queue Time with (9, 20, 18, 6, 16) = 69 instances


Fig. 8 – Resource usage with (9, 20, 18, 6, 16) = 69 instances


Fig. 9 – Completion Time with (9, 20, 18, 6, 16) = 69 instances

• Compare the parameters sensitivity, i.e., the outcome variation with respect to the

unit variation of the parameter. Starting from (9, 20, 18, 6, 16), decrease a resource

by one unit and assess the total duration.


Resources Total Duration Difference w.r.t 3:11:59 (192m)

(8, 20, 18, 6, 16) 3:24:01 +12:02 => NO (more than 200m)

(9, 19, 18, 6, 16) 3:22:05 +10:06 => NO “

(9, 20, 17, 6, 16) 3:13:05 +01:06 => YES (less than 200m)

(9, 20, 18, 5, 16) 3:16:12 +04:13 => YES “

(9, 20, 18, 6, 15) 3:12:36 +00:37 => YES “

“large and complex organizations are a tangible manifestation of advanced technology ......

Documents