A Model-based Software Technology Proposal
Vahur Kotkas, Jaan Penjam, Ahto Kalja and Enn Tyugu
Institute of Cybernetics at Tallinn University of Technology, Akadeemia 21, Tallinn, 12618 Estonia
Keywords: Model-based Software Development, Application Engineering.
Abstract: In this paper we propose a technology for model-based software development. The technology separates
domain engineering and application engineering and automates the actual executable code generation. This
approach has been used extensively for simulations and we believe it is also applicable for more general
software development.
1 INTRODUCTION
Model-based software development is a way to
overcome the increasing complexity of software
products and their changeability (Vitkin et al.,
2006). It is based on dividing the software
development into two separated processes: domain
engineering and application engineering. Both
include software development as a part. The first
process provides software assets for the use in the
second process. Software assets are the reusable
resources used in application engineering. Examples
of software assets include domain models, software
architectures, design standards, communication
protocols, code components and application
generators. This facilitates software development by
raising the conceptual level of application
programming.
2 MODEL-BASED SOFTWARE
DEVELOPMENT
APPROACHES
The idea of model-based software development is
not new. It has been around for almost 30 years, but
has not become a widely accepted paradigm. Its
most successful applications are in simulation
software, there are well known specialized products
like Simulink (Dabney and Harman, 1997) or Scicos
(Scicos Wiki). One continuous effort in this field is
pursued in NASA (Cooke et al., 2006). Aerospace
applications, including software for the International
Space Station (ISS), use model-based development
extensively. NASA puts strict requirements on the
model-based software development. As the authors
say, the production-quality program synthesis is the
keystone for full-cycle model-based programming.
Without this capability, model-based programming
is limited to being a prototyping tool whose utility
ends after detailed design, when the production code
is developed manually. We completely agree with
this statement, and take it into account in the
development of our methods.
Considerable amount of work is being done in
improving the existing UML-based approaches with
the aim of providing automated support to the
software development (Engels et al., 2007) and
language development (Selic, 2007). This approach
is also related to Model Driven Architecture (MDA)
advocated by the OMG (Object Management Group)
and to the development of domain specific
languages (DSL), (see van Deursen and Klint, 2001;
or Mernik et al., 2005), because they all have the
development of user friendly and automated
problem solving tools as a goal. This approach
includes the usage of UML-based models and
metamodels. It concentrates either on the research of
transformation rules for transforming an initial
specification (a model) into another model or an
executable code (Whittle, 2002), or on the
development of rules that represent the operational
semantics (Engels et al., 2007) or even immediately
perform the required computations. Despite its
popularity and numerous research papers, this
approach has remained rather theoretical. Sound
criticism on this approach can be found in (de Niz,
2007; Rath, 2006). In particular, the universal
character of UML and its orientation at software
312
Kotkas V., Penjam J., Kalja A. and Tyugu E..
A Model-based Software Technology Proposal.
DOI: 10.5220/0004348203120315
In Proceedings of the 1st International Conference on Model-Driven Engineering and Software Development (MODELSWARD-2013), pages 312-315
ISBN: 978-989-8565-42-6
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
implementation is an obstacle for its usage in
domain modeling where the high-level domain
specific concepts must be handled.
Another research direction in the model-based
software development is the direct usage of
graphical tools (without transformation of DSL or
source models into the UML form) (Sprinkle and
Karsai, 2004), e.g. the MetaEdit+ (see MetaCase,
2012; or Tolvanen and Kelly, 2009), but no
universal tools for high-level domain modeling exist
yet. One loses the rich collection of UML-based
presentations in this case, but gets more freedom in
developing the automation methods.
One more approach to model-driven software
development has been made by the Eclipse
community. Eclipse Modeling Project (EMP)
(Gronback, 2009) that includes Eclipse Modeling
Framework (EMF), Graphical Modeling Framework
(GMF) and the Generative Modeling Tools (GMT)
is a relatively new collection of technologies for
building DSLs. EMF provides a basis for abstract
syntax development (Encore model) which
corresponds (is mapped) to a textual concrete syntax
or a graphical concrete syntax. EMP includes
various components, such as UML2, OCL, QVT,
EMOF, XMI, etc. that enable users to develop DSLs
using MDA. Generally speaking, EMP is a powerful
tool, but it requires a lot of effort to develop a
working DSL from scratch.
The important topic relevant to our technology is
a generative programming paradigm. It is about
manufacturing software products out of components
in an automated way (Czarnecki and Eisenecker,
2000). This definition precisely fits into our model-
based technology proposal. To go into more details,
the generative programming focuses on software
system families rather than one-of-a-kind systems.
Such family members can be automatically
generated based on a common generative domain
model that includes implementation components and
the configuration knowledge mapping between a
specification and a finished system or component. A
good source of information that addresses the issues
and presents tools and applications of the generative
programming is the GPCE international conference
(see GPCE proceedings).
(Grigorenko et al., 2005-1; 2005-2; 2006)
describes a visual tool that can be used for
specifying models. In this sense, it supports the
direct usage of graphical tools, and does not require
transformation of DSL or source models into the
UML form. The technology has similarities with the
NASA approach, but relies on other program
construction methods.
Positive experiences are recieved with applications
of program synthesis in cyber defence and
information assurance (Kivimaa et al., 2008-2009),
in simulation (Grossschmidt and Harf, 2008, 2009;
and Ojamaa, 2008), as well as in composition of
services on large service models (Maigre, 2008).
3 GOALS AND BASIC
HYPOTHESES OF THE
PROPOSED APPROACH
The approach is intended for approbation of new
ideas in software engineering – a model-based
software technology that is based on automatic code
construction by means of logical and planning
methods.
The proposed model-based technology will
provide high degree of automation based on
application of logical methods of program
construction. A workflow of the software
development according to this technology is
presented on Figure 1.
The domain engineering is the process where
domain experts develop/obtain ontology (1) for a
software system.
Software developers implement this ontology by
developing classes (2), visual classes and
metaclasses (4). Metaclasses are code components
(classes are basic building blocks to be included in
the developed software) extended with specifications
that enable their automated handling during the steps
(7), (8) and (9). Visual classes are metaclasses with
visual representation to be used in step (6) for
creating schemes.
The application engineering consists of
requirements development (5) for a particular
software product and of specification writing (6) in a
precise language that is an input language of the
automation tool.
The steps (7, 8, 9) from specification to final
code are performed automatically (see Matskin and
Tyugu, 2001 for an idea).
This is the scheme tested already (also in large-
scale) for complex simulations (Grossschmidt 2008,
2009; Maigre, 2008; Harf and Grossschmidt, 2012).
The main hypothesis is that instead of
transforming specifications, first, from domain
specific notations into UML, and thereafter using
UML tools, as is the mainstream of model-based
software development (and used in the model-driven
architecture – MDA), one can transform the
specifications directly into logic, and use logical
AModel-basedSoftwareTechnologyProposal
313
Figure 1: Software development workflow
tools for representation of semantics as well as for
automatic construction of executable code. This
hypothesis is supported by experience in automated
program construction (Matskin and Tyugu
, 2001)
and, more generally, by transformation of high-level
and visual specifications into executable code
(Grigorenko and Tyugu, 2010; Tyugu and Valt,
1997). UML can be used in a conventional way for
requirements specification and development of
classes even in this case. However, the real success
of the technology can be achieved when the logic of
synthesis is extended and more expressive
specifications, than used today in the model-based
software development, will be available. We foresee
the need for branching, resource consumption and
distribution specification support, e.g. extension
with linear-logic.
4 CONCLUSIONS
An approach of model-based software development
technique is presented. Our belief is that such
technique becomes applicable in practice only when
automatic code construction by means of logical and
planning methods is applied. For that the
development has to be separated into several levels
of engineering where domain engineering and
application engineering is separated. The proposed
approach needs more effort being put in domain
engineering and application engineering while
developing the executable code is fully automated.
ACKNOWLEDGEMENTS
This research was partially supported by the target-
financed theme No. SF0140007s12 of the Ministry
of Education and Research as well as by the ERDF
funded Centre of Excellence in Computer Science.
REFERENCES
Vitkin, L., Dong, S., Searcy, R. and Manjunath, B., 2006.
Effort Estimation in Model-Based Software
Development. 2006 SAE World Congress Detroit,
Michigan.
Dabney, J. B., Harman, T. L., 1997. Mastering
SIMULINK. Prentice Hall PTR.
Scicos Wiki, http://en.wikipedia.org/wiki/Scicos
Cooke, D. E., Barry, M., Lowry, M., Green, C., 2006.
NASA’s Exploration Agenda and Capability
Engineering.
Engels, G., Soltenborn, C., Wehrheim, H., 2007. Analysis
of UML Activities Using Dynamic Meta Modeling. In:
Formal Methods for Open Object-Based Distributed
Systems, 9th IFIP WG 6.1 International Conference,
FMOODS 2007, Paphos, Cyprus, June 6-8, 2007, pp.
76-90.
Selic, B., 2007. A Systematic Approach to Domain-
Specific Language Design Using UML. In Proceedings
of the 10th IEEE international Symposium on Object
and Component-Oriented Real-Time Distributed
Computing (May 07 - 09, 2007). ISORC. IEEE
Computer Society, Washington, DC, 2-9. 2007.
Object Management Group. Model Driven Architecture.
http://www.omg.org/mda.
MODELSWARD2013-InternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment
314
van Deursen, A., Klint, P., 2001. Domain-specific
language design requires feature descriptions. Journal
of Computing and Information Technology. Vol. 10.
Mernik, M., Heering, J., Sloane, A., 2005. When and how
to develop domain-specific languages. ACM
Computing Surveys (CSUR) Volume 37, Issue 4.
Whittle, J., 2002. Transformations and software modeling
languages: Automating transformations in UML.
LNCS, vol. 2460. Springer-Verlag
de Niz, D., 2007. Diagram and Language for Model-
Based Software Engineering of Embedded Systems:
UML and AADL. Software Engineering Institute white
paper, Carnegie Mellon University.
Rath, I., 2006. Declarative Specification of Domain
Specific Visual Languages. Master’s thesis. Budapest.
Sprinkle, J., Karsai, G., 2004. A domain-specific visual
language for domain model evolution. Journal of
Visual Languages and Computing, Vol 15 (3-4),
Elsevier.
MetaCase, 2012. http://www.metacase.com.
Tolvanen, J.-P., Kelly, S., 2009. Metaedit+: defining and
using integrated domain-specific modeling languages.
OOPSLA 2009 Companion, pages 819–820. ACM.
Gronback, R., 2009. Eclipse Modeling Project: A Domain-
Specific Language (DSL) Toolkit. Addison-Wesley
Professional.
GPCE proceedings. http://www.informatik.uni-trier.de/
~ley/db/conf/gpce/index.html.
Grigorenko, P., Saabas, A., Tyugu, E., 2005. Visual tool
for generative programming. ACM SIGSOFT
Software Engineering Notes, 30, 5, 249-252.
Grigorenko, P., Saabas, A., Tyugu, E., 2005. COCOVILA
– Compiler-Compiler for Visual Languages. In: J.
Boyland, G. Hedin. Fifth Workshop on Language
Descriptions Tools and Applications LDTA2005.
ETAPS, p. 101 – 105.
Grigorenko, P., Tyugu, E., 2006. Deep Semantics of
Visual Languages. In: E. Tyugu, T. Yamaguchi (eds.)
Knowledge-Based Software Engineering. Frontiers in
Artificial Intelligence and Applications, vol. 140. IOS
Press, p. 83 - 95.
Kivimaa, J., Ojamaa, A., Tyugu, E., 2009. Graded
Security Expert System. Proc. CRITIS08, LNCS 5508
Springer, 279-286.
Kivimaa, J., Ojamaa, A., Tyugu, E., 2009. Managing
Evolving Security Situations. MILCOM 2009:
Unclassified Proceedings, October 18-21, 2009,
Boston, MA. Piscataway, NJ: IEEE, 1-7.
Kivimaa, J., Ojamaa, A., Tyugu, E., 2008. Pareto-Optimal
Situation Analysis for Selection of Security Measures.
Proc. MilCom 2008, 7 p.
Grossschmidt, G., Harf, M., 2008. Modelling and
simulation of fluid power systems in an intelligent
programming environment. Proc ISC'2008 : June 9-11,
2008, Lyon, France, Proceedings: Ostend: EUROSIS,
2008, (A publication of EUROSIS-ETI), 224 - 230.
Grossschmidt, G.; Harf, M., 2009. COCO-SIM - Object-
oriented Multi-pole Modelling and Simulation
Environment for Fluid Power Systems.
Part 1:
Fundamentals. International Journal of Fluid Power,
Vol. 10, No. 2, 2009, pp. 91 - 100. Part 2: Modelling
and simulation of hydraulic-mechanical load-sensing
system. International Journal of Fluid Power, Vol. 10,
No. 3, 2009, pp. 71 - 85.
Harf, M.; Grossschmidt, G., 2012. Modeling and
simulation of an electro-hydraulic servovalve in an
intelligent programming environment. In: ASME 2012
11th Biennal Conference on Engineering Systems
Design and Analysis (ESDA 2012), July 2-4, 2012,
Nantes, France, [Proceedings]; New York, ASME,
2012, 1-9.
Ojamaa, A., Tyugu, E., 2008. Rich Components of
Extendable Simulation Platform. Proc.
WORLDKOMP'07: MSV2007, CSREA Press, 2007,
p. 121 - 127. M. Virvou, T. Nakamura (eds.)
Knowledge-Based Software Engineering. Proc. 8th
JCKBSE. IOS Press, p. 49 – 58.
Maigre, R., Grigorenko, P., Küngas, P., Tyugu, E., 2008.
Stratified Composition of Web Services. In: M.
Virvou, T. Nakamura (eds.) Knowledge-Based
Software Engineering. Proc. 8th JCKBSE. IOS Press,
p. 49 – 58.
Maigre, R., Küngas, P., Matskin, M., Tyugu, E., 2008.
Handling Large Web Services Models in a Federated
Governmental Information System. Proc. 3-rd
International Conference on Internet and Web
Applications and Services. IEEE Computer Society &
CPS, p. 626 – 631.
Matskin, M, Tyugu, E., 2001. Strategies of Structural
Synthesis of Programs and Its Extensions. Computing
and Informatics. v.20, p.1 -25.
Grigorenko, P., Tyugu, E., 2010. Higher-Order Attribute
Semantics of Flat Declarative Languages. Computing
and Informatics.
Tyugu, E., Valt, R., 1997. Visual programming in NUT.
Journal of visual languages and programming, v. 8,
pp. 523 – 544.
Czarnecki, K., Eisenecker, U. W., 2000. Generative
Programming: Methods, tools, applications, Boston.
AModel-basedSoftwareTechnologyProposal
315