S-CUBE LP: Service Network Analysis & Prediction Tool (SNAPT)

S-Cube Learning Package
Service Networks Visualization:
Service Network Analysis & Prediction Tool
(SNAPT)

University of Crete (UoC)
Mariana Karmazi, Christos Nikolaou, Pantelis Petridis, George
Stratakis

www.s-cube-network.eu

Learning Package Categorization

S-Cube

Business Process Management

(Performance) Analysis and Design of Service Networks

Service Network Analysis and Prediction Tool (SNAPT)

Learning Package: Service Network Analysis and Prediction Tool © S-Cube - 2

Learning Package Overview

 Problem Description
 Service Network Analysis and Prediction Tool (SNAPT)
– SNAPT Overview and Fundamental Concepts
– SNAPT Metamodel and Visualization Techniques
– From Service Network Models to initial draft Business Process Models
and simulation models

 Summary


Background: Service Systems (or
Service Networks)
 Service system: dynamic co-creation configuration of resources (people,
organizations, shared information) and technology, connected together
through value propositions (Spohrer, Maglio)
– Proposed in order to model, analyze and optimize interactions among various
network partners.
– High level of abstraction, hiding details regarding concrete interactions in
terms of business processes
- Model services that are offered and consumed by business entities
- Service providers (providing a set of service offers)
 Formatted mainly because of: globalization, advances in ICT, pressure for
innovation, increased competition, constant change of customers’
demands, which lead to increased focus on core competencies (or
strengths) and outsourcing.
 Based on a new marketing discipline: Service-Dominant (S-D) Logic
 Service networks are considered as projections of service systems and
they are embedded in Service Ecosystems

Background: Service Ecosystem

 A Service Ecosystem is a socio-technical environment consisting of:
– All the services available in a particular sector of the economy (e.g. home
electronics, online media, etc.)
– All the supporting (enabling) good and services (e.g. Banking, building
mainetance, power and telecom utilities, brokers, distributors, etc.)
– All the regulating and supervising authorities


Jim Spohrer (IBM): Multiple Approaches
to Study Service Systems


Perspectives on Service Systems
Modeling and Analysis
 Business Perspective (economic and marketing
viewpoint): Conceptual modeling and analysis techniques
studying service networks in a high abstraction layer depicting
the entities participating in the network while analyzing
network vitality and calculating value created for each
participant and for the network as a whole.
– Value chain, Value Networks

 IT Perspective: Deals with the alignment and coordination of
the participating entities’ business processes and information
systems in order to achieve the agreed-upon business
outcome
– Business Process Management (BPM) and its lifecycle
– Service-oriented architecture (SOA)


Problem Description

 Thus, there is a need for a holistic approach combining the concepts
underpinning BPM and SOA in order to support service systems prevailing
in the networked economy
– Propose a unified modeling methodology combining concepts from the
business perspective and the IT perspective
– Target both
- Business analysts
- IT experts
 SNAPT Vision:
– Visualize service networks (SNs), define business metrics and goals to SNs in
terms of Key Performance Objectives (KPOs), monitoring of KPIs
corresponding to KPOs sets, take corrective actions (e.g., violations)
– A prototype tool for constructing service network models and transforming into
initial business process models based on BPMN, bridging the gap between
tools and concepts.




 Summary


Overview of SNAPT

 SNAPT is a prototype tool for:
– Modeling service networks as a set of services and business entities based on a
proposed meta-model
– Adding Key Performance Objectives (KPOs) to services within SNs
– Analyzing performance of service networks
– Extracting draft business process models out of service networks models
- BPMN 1.2 support, Eclipse BPMN editor
- IBM Websphere Studio process diagrams
– Support for simulation tools (e.g., Vensim, iThink)
- Currently, service networks models are
mapped to simulation models
supported in Vensim


SNAPT Architecture in brief

 SNAPT has been developed following the Model-Driven Architecture
(MDA)
 Eclipse platform has been utilized for the development process taking
advantage of the plug-ins extension mechanisms in order to allow future
extensions or modifications to the tool.


SNAPT Models and Plug-ins


SNAPT Fundamental Concepts

 Foundation principle: any business can be modeled as a
service.
– Products as a service = it is the delivery of the product that comprises
the service offered to the end customer
– Consistent with Service-Dominant (S-D) Logic

 The existence of a service network implies that there is a
single service or a bundle of services that a key business
entity delivers to an end customer.
– In the car industry domain, the process of manufacturing a car can be
modeled as a service network encompassing all the entities working
together to deliver a car
- The service network provides a single service: the car


SNAPT Fundamental Concepts

 A Service Network is defined as a set of business entities and
services and it can be visualized as a graph of nodes
– Nodes correspond to business entities
– Arcs correspond to services offered and consumed by the business
entities inside the network
- An arc implies an economic exchange
- Origin point of the arc reveals the business entity that offers the
service
- The end point of the arc reveals the business entity that consumes
the service




 Summary


SNAPT Service Network Metamodel


Metamodel Concepts

 Business Entity: any independent economic entity that provides and/or
consumes services in a service network
– Business entities may offer various distinct and even unlike services at the
same time, cooperating with a lots of independent business entities.
– When modeling a service network, only those services that contribute to the
final service offering is modeled for each business entity
 Business entities generate value from their participation in the network
 Each business entity is assigned to a role in a particular network
 Four types have been distinguished based on the functional properties:
– End-customers
– Enablers
– Service Sub-Network
– Participants


A Business Entity/Service Provider


Metamodel Concepts

 End Customer: only consumes service provided by the
network
– They do not contribute to the service composition
– They actually offer a service to the network by providing feedback, e.g.
feedback concerning their experiences

 Enabler: offer a service that enables the delivery of other
services
– Always interact with both service provider and service consumer
– For example: FedEx, intermediate payment services like Paypal or
Google CheckOut.


Metamodel Concepts

 Service Sub-Networks: they have an internal structure of their
own and nestle an entire service network that provides and/or
consumes services.
– Restriction: each service offered or consumed by a service sub-
network must also be offered or consumed by a single business entity
inside the sub-network

 Participants: an ad-hoc business entity which usually used to
refer to the service providers.


Metamodel Concepts

 Services: refer to both goods and services, tangible and
intangible in nature. Denote what is exchanged in the network
– Connects business entities in 1-to-1 relationships, ‘offer’ and
‘consume’
– ‘Offering’ is represented with a solid-line arrow originated from the
business entity node which acts as service provider
– ‘Consumption’ is depicted as a dashed-line arc in the opposite
direction originating from the business entity node that consumes the
service – service consumers

 “Enablement Service”: connects an enabler with another
business entity or directly with the service it enables


Metamodel Concepts

 Key Performance Objectives (KPOs): used to model business and
performance objectives
– Associate a business metric with a service in the network and describe the
expected performance of the underlying business processes from both the
source and target business entities
– Reflect the expected target value as declared by a business analyst
– For any given service, a service provider has his own business goals reflected
to the KPOs that he will try to satisfy. At the same time, the consumer has
some requirements that the service must meet and these should also be
reflected to the KPO Model. As a result, in our meta-model service offerings
are related to KPOs, and so do service consumptions
– Key Performance Indicators (KPIs): are business metrics used on the
Business Process Management layer for as a part of the monitoring process
for measuring the performance of business processes
- KPIs contains the measured value of a business metric in contrast to the
expected value declared by a KPO


Snapshots: Visualize Service Networks

Participant offers a Single service to the End Customer

Enablement Service: an Enabler enables the delivery of the Service 1


Snapshots: Service Sub-Networks

Service Network ServiceSubNetwork
Sub-network input

Sub-network output

Service “Supplies” consumed by “Service” offered by
ServiceSubNetwork in the SN (left), is ServiceSubNetwork in the Service
mapped to an input port in the Network (left), is mapped to an output
ServiceSubNetwork (right). port in the ServiceSubNetwork (right).


Snapshot: Assign KPOs

 SNAPT provides a KPI Library based on
the APQC Process Classification
Frameworks
 Fully compatible with IBM
Websphere Business Modeler
 SNAPT updates its internal KPI library
from the KPIs Repository
 REST-based interface

 SNAPT user can select from the library
the desired KPO to assign to a service


Case study: Car Repair Service Network

 The purpose of this network is to efficiently deliver to the car owners the
service of “Parts and Repair”
 In order for the Dealers to deliver the “Parts and Repair” service to the
Car Owners, they must first order the parts with the help of the Parts
Manager and then consume one of the “Parts” service delivered by Car
OEM or the Third Party Suppliers, together with the “Repair” service
provided by the technicians and taking into account the “Advice for
Repairs” service delivered by the CAR OEM. The CAR OEM delivers the
“Parts” and the “Advice for Repairs” service after consuming the
corresponding services from the Supply Chain Supplier and the Help Desk
Experts, respectively.


Model



– From Service Network Models to initial draft Business Process
Models and simulation models

 Summary


From Service Network Models to initial
draft Business Process Models
 SNAPT supports a methodology for mapping Service Network
Models to Business Process Models
 Two sets of transformation rules are proposed and supported
by SNAPT
– The 1st set maps SN models to collaborative business process models
according to BPMN v.1.2 standard and the export format is compatible
with the Eclipse BPMN Editor, an open source business process
diagram editor
– The 2nd set maps SN models to process models based on the format
supported by IBM WebSphere Business Modeler Advanced.
- A commercial business process modeling and analysis tool


Sequencing of Services (1/3)

 The delivery of the service offered by a service network
implies that the resources and back-end systems of the
business entities are integrated and coordinated accordingly
in order to achieve connection of entire business value chains
that will deliver the desired outcome.
 However, service networks models are highly abstract in
nature and they do not include any operational details, like
sequencing of processes, message exchanges, etc.
 So, it’s mandatory to properly annotate services in the
service network models to define the order of services in a
service network model
– Identify composite services


Sequencing of Services (2/3)

 Sequencing information of each service s is created relatively to the
set of services that are offered to the source business Entity of
service s
 Gateways are used to express sequencing, which can be nested in
any order
– Sequential Block: this gateway implies that any of its children elements is
delivered in series, one after the other
– AND Block: this gateway specifies that its elements should be delivered in
parallel
– XOR Block:denotes that exclusively one of the elements in the block must be
delivered.

 To sum up, any sequencing service (a service with sequencing
information attached) is decomposed to several services that will be
delivered in the order defined by gateways


Sequencing of Services (3/3): Simple
Sequential Block Example
 Example of Sequential Block: In order for Service1 to be delivered to
Participant2, both Service2 and Service3 must first be delivered to Participant1 in
order.
Annotation indicating
Sequencing Services

Sequence Order


From SNAPT to Eclipse BMPN Editor

Mapping from SN constructs
to BPMN elements

Mapping a single service
delivery to a generic
BPMN workflow



 The transformation process is more complex if the service
network is annotated with sequencing information
– The service’s sequencing information is mapped to both generic and
complex workflows, depending on the service or SequencingService
objects composed of.
– These workflows are processed in a specific manner, depending on
the sequencing constructs used in the sequencing information.
– Objects contained in
- … a SequentialBlock, are mapped to workflows that are processed
in series in the order indicated by the Sequence, XORSequence or
ANDSequence objects.
- … an ANDBlock are mapped to workflows that are connected via a
BPMN parallel gateway,
- … a XORBlock are mapped to workflows that are connected via a
BPMN exclusive data-based gateway

Example
 Based on the Simple Sequential Block Example


From SNAPT to IBM WebSphere
Business Modeler

Mapping from SN
constructs to IBM
Modeler elements

Mapping a single service delivery to a generic workflow


From SNAPT to IBM WebSphere
Business Modeler


to Eclipse BPMN diagram


to IBM WebSphere Modeler

1
(a)

(b)

(c)

(a)
2
(b)

(c)


From SNAPT to VENSIM tool

 A system dynamics model in Vensim tool also consists of
variables and arrows that represent the relations and
specifically the dependencies among the variables.
 Business entities are mapped to either a constant or an
auxiliary variable
 Services are mapped to variables, as well. For each Service
Network Model Service, three variables are declared
corresponding to the two business entities and the service;
two arrows connect the service variable to the source and
target business entity


From SNAPT to VENSIM models




 Summary


Summary

 Towards bridging the world of business analysts and IT
experts including the concept of service systems
 SNAPT serves as a hub providing appropriate outputs to both
simulation tools that analyze the vitality of these networks as
well as to BPM suites, for supporting the underlying business
processes which connect the systems of the involved
participants.
– To this extent SNAPT models are transformed and extracted into a
draft form of collaborative business processes based on the BPMN
format and the IBM’s WebSphere Business Modeler business process
model format.


Further Reading

P. Petridis, C.Nikolaou:Towards a universal Service Network-centric framework to
design, implement and monitor Services in complex Service Ecosystems: The
Service Network Analysis & Prediction Tool (SNAPT). Department of Computer
Science, University of Crete, Heraklion, Master Thesis (Msc) 2010.

G. Stratakis, C. Nikolaou: Analyzing Service Networks from different
perspectives usingthe Service Network Analysis & Prediction Tool (SNAPT)
Department of Computer Science, University of Crete, Heraklion, Master Thesis
2011.


Acknowledgements

The research leading to these results has
received funding from the European
Community’s Seventh Framework
Programme [FP7/2007-2013] under grant
agreement 215483 (S-Cube).


S-CUBE LP: Service Network Analysis & Prediction Tool (SNAPT)

More Related Content

Similar to S-CUBE LP: Service Network Analysis & Prediction Tool (SNAPT) (20)

More from virtual-campus (20)

Recently uploaded (20)

S-CUBE LP: Service Network Analysis & Prediction Tool (SNAPT)