semantic web technologies in model-driven control application engineering · pdf...
TRANSCRIPT
Semantic Web Technologies
in Model-driven Control
Application Engineering
Teaching associate David Hästbacka, professor Seppo Kuikka
Research area: IT Architectures for Industrial Control
Department of Automation Science and Engineering
TUT W3C Web Technology Day: Linked Data for Science and Industry
October 3, 2012
Contents
• Introduction and Background
• Metamodeling and Ontology Modeling
• Ontology Semantics in Control Application Models
• Ongoing Research
• Conclusion
5.10.2012 2
Introduction and Background
• Model-driven engineering (MDE) and model-driven
development (MDD) promote models as primary
engineering artefacts
• In our previous work a model-driven development
process ranging from process design to executable
control applications has been developed
5.10.2012 3
Hästbacka, D., Vepsäläinen, T., Kuikka, S.:
Model-driven Development of Industrial Process Control Applications.
Journal of Systems and Software 84 (2011) 1100-1113 Doi:10.1016/j.jss.2011.01.063
Ontology Semantics in
Control Application Models
• Ontologies can be used in software engineering to provide
logic-based formalisms and semantics to concepts
• In the approach OWL ontologies are used as a
supplement to (UML Automation Profile based) models
– A semantic layer on top of the modeling language
– Support development of control software applications in
• classification of concepts and structures
• describing and linking other information
• analysis of constructs
5.10.2012 4
From UML AP Models to
OWL Semantics
Ontology individuals
Domain ontology
Model instances
MDE/MDD
Environment
Model package
UML AP
metamodel
Metamodel
mapping
Implemented once,
depends on the
profile (metamodel)
version being used.
Instance model
generation
Performed on-line
for models being
designed.
UML AP ontology
(rdf/owl)
UML AP ontology
(rdf/owl)
Reasoning
and analysis
conforms to conforms to
Domain ontology
(concepts and
knowledge of the
particual domain)
Ontology individuals
(model instances,
data)
5.10.2012 5
Reasoning Example
6 *Oma nimi ja esityksen aihe vaihdettava alatunnisteeseen
Reasoning
example:
Concept
generalization
6
Utilization in an IDE
5.10.2012 7
Knowledge base
Domain
knowledge
Model
instance
knowledge
applied to
applied to
Use case
specifc
knowledge
• Make the engineering environment more intelligent and
aware of what is being developed
• Enable existing knowledge to be integrated to the
engineering context to support work tasks
OWL model, events
Transformer
Semantic KB(of model)
UML AP Tool Editor Contextual UI Views
UMLAPto OWL
UMLAP model,events
Ongoing Research (1/2)
• Describe design patterns using
ontologies to be used in
industrial control application
development
• Provide sophisticated guidance
for choosing design patterns
that promote desirable features
5.10.2012 8
UML AP Tool Editor
Design Pattern Assistant View
Design Pattern Semantics KB
Search request
Semantic KB (of Model)
Discipline
Safety system
engineering
Domain (within the
discipline)
Scope (the features
that implement the
pattern (may relate to
paradigm))
Software Hardware
Granularity (the level
at which the pattern
addresses the
system)
System
architecture
Design
Requirements
engineering
Communi-
cationDevelopment
process
Paradigm (the
paradigm the pattern
utilizes, the way the
safety systems are
described)
Purpose (the kind of
problem the pattern
solves)
Risk
evaluation
System
analysis
Hazard
analysis
Context
analysis
Requirement
gathering
DocumentationSafety
priorization
System safetyDistribution
Fault-tolerance
Design
principle
System
operabilityFault detection
Data loggingStandard
compliance
Development
tools
Traceability
Process
patterns
COTS
Dependability
Non-blockingInformation
sharing
IEC 61508
Function block
Process
automation
Machinery
automation
Batch
automation
E/E/PE
systems
Passive safety
systems
Training Manuals
WarningsBlocking
Idiom
High demand
Low demand
Software
architecture
System
developent
Application
software
Safety-related
hardware
Firmware
Co-operation
Communi-
cation
Application
domain
Problem
relatedBasic principle
• Work support for control application engineers
– Utilize semantic modeling context
– Provide related engineering information that assists the
development
– Improve sharing of knowledge in engineering teams
Ongoing Research (2/2)
5.10.2012 9
TransformerSemantic KB(of Model)
UML AP Tool Editor Work Support View
Engineering information,
manuals, handbooks
Annotations, comments
Conclusion
• Ontology semantics have been successfully used as a
supplement to control software modeling
• Classification and inference of models enables many
opportunities to support engineering
• Identify structures outside the scope of the metamodel
• Include reusable design information for the given design context
• Provide interoperability between concepts of different modeling
methods
5.10.2012 10
References
• Hästbacka, D., Kuikka, S. (2012) Semantics Enhanced Engineering and Model Reasoning for
Control Application Development, MTAP, Vol. xx, No. x, 2012. DOI: 10.1007/s11042-012-1134-9
(In press)
• Hästbacka, D., Kuikka, S. (2012) Semantics and Reasoning for Control Application Engineering
Models, LNCS, Vol. 7267, pp. 647-655. DOI: 10.1007/978-3-642-29347-4_75
• Hästbacka, D., Kuikka, S. (2011) Bridging UML Profile Based Models and OWL Ontologies in
Model-driven Development – Industrial Control Application. International Workshop on Future
Trends of Model-Driven Development (FTMDD 2011), Beijing, China, June 2011, pp. 13-23. DOI:
10.5220/0003561900130023
• Hästbacka, D., Vepsäläinen, T., Kuikka, S. (2011) Model-driven Development of Industrial Process
Control Applications. Journal of Systems and Software, vol. 84, no. 7, pp. 1100-1113. DOI:
10.1016/j.jss.2011.01.063
5.10.2012 12