SIMULTECH 2014

Supply Chains Modelling and Simulation Framework:

Graph-Driven Approach Using Ontology-Based
Semantic Networks and Graph Database

Mahmoud Elbattah, Owen Molloy

National University of Ireland Galway
m.elbattah1@nuigalway.ie
SIMULTECH 2014

Outline

• Supply Chains Background

• Research Problem

• The Proposed Framework & Methodology

• Expected Outcomes

2
SIMULTECH 2014

Supply Chains Background

3
SIMULTECH 2014

What is a Supply Chain?

Product Flow

.
. Supplier Manufacturer Distribution Centre Retailer Customer
.
Information Flow

. Manufacturer Distribution Centre Retailer Customer

Supplier
.
.

Supplier Manufacturer Distribution Centre Retailer Customer

Supply Chains Background 4

SIMULTECH 2014

Research Problem

5
SIMULTECH 2014

Outline of Challenges

Conceptual •How should supply chain

Modelling knowledge be represented?

•How to build models that

Challenges of Model facilitate sharing of supply
Modelling & Share-ability chain knowledge?
Simulation of Supply
Chains

Simulation •Building simulation models

Tools require much time, effort and cost.
Usability

Model
Scalability •Rapidly growing supply chains imply
scalability issues. Research Problem 6
SIMULTECH 2014

Potential Gaps in Literature

• The lack of a standard basis for modelling supply chains.
• The developed models might not help with scalability for large-scale
supply chains.
Conceptual
Modelling • The Lack of recognition that supply chains are neither completely
discrete nor continuous, but a mixture of both.

• Apart from (Fayez, Rabelo, 2005), the ontology mainly addressed the
Supply strategic level of supply chains.
Chains • The shortage of industry-specific ontologies.
Ontology

• Simulation tools were mostly convenient for simulation experts.

• Automatic generation (model-driven architecture) of simulation models
Simulation
has been little addressed.
Tools

Research Problem 7
SIMULTECH 2014

The Proposed Framework

8
SIMULTECH 2014

Outline of Objectives
• Providing a semantic-based modelling method for supply chains.
• Investigating the flexibility and scalability provided by graph database
Conceptual for building complex large-scale supply chain models.
Modelling

• Developing generic ontology that can describe the strategic, tactical

Supply and operational levels of supply chains.
Chains • Developing specific ontology for healthcare supply chains.
Ontology

• Automatic generation of simulation models based on high-level

Simulation conceptual models to help non-simulation experts.
Models

The Proposed Framework 9

SIMULTECH 2014

The Proposed Framework Overview

Semantic Generic Domain

Networks Ontology Ontologies
Conceptual
Modelling

Graph Database

Simulation
Simulation Model
Software
Auto-Generator
(AnyLogic)

The Proposed Framework 10

SIMULTECH 2014

Conceptual Modelling Approach

11
SIMULTECH 2014

Supply Chains as Big Graphs

A typical supply chain example.

• A virtual complex network of suppliers, manufacturers, wholesaler, retailers

and customers connected through upstream and downstream linkages.

• Apparently, it can be conceivable to consider modelling supply chains as

constructing “Big Graphs”.

Conceptual Modelling Approach 12

SIMULTECH 2014

Supply Chains as Semantic Networks

An example of a semantic network.

Nodes  The supply chain participants (entities) interconnected.

Arcs  Predicates that can represent properties or relationships.

Conceptual Modelling Approach 13

SIMULTECH 2014

Supply Chains Ontology

Generic •Based on the SCOR model.

Ontologies •Classifying discrete & continuous
attributes

Extraction of
Ontology

Domain •Industry-specific ontology for

Ontologies healthcare.

Conceptual Modelling Approach 14

SIMULTECH 2014

Models and Ontologies Storage with

Graph Database
Supply •Flexible graph-based models.
Chain
•Scalable schema-less models.
Models

Graph Database

•Typically faster for associative

Supply
datasets.
Chain
Ontologies •Potential flexibility to align
the ontology with RDF-based
format.

Conceptual Modelling Approach 15

SIMULTECH 2014

Building Simulation Models

16
SIMULTECH 2014

Automatic Generation of Simulation Models

Supply Chain
Ontology

Conceptual Automatic Simulation

Scenario Details
Model Generator Model

Simulation
Model
Parameters

17
SIMULTECH 2014

Expected Outcomes

18
SIMULTECH 2014

Expected Outcomes
• Higher flexibility and share-ability of supply chain models through
semantic graph-driven models.

• Extended potentials for storing large-scale supply chain models using

graph database.

• Extracting generic ontology for supply chains.

• Developing specific ontology for healthcare supply chains.

Expected Outcomes 19
SIMULTECH 2014

Expected Outcomes (cont’d)

• Capability to build combined discrete-continuous models based on
discrete-continuous classification of ontology attributes.

• Helping non-simulation experts by automatic generation of simulation

models.

Expected Outcomes 20
SIMULTECH 2014

Thank You!
Mahmoud Elbattah
m.elbattah1@nuigalway.ie

21