![](bg8.png)
count such as the ability to exchange models, generate
documents, support collaborative work, compliance
with certification, etc. It is difficult to rely on existing
surveys as they are usually quickly outdated given the
constant evolution of the tools, acquisition between
vendors and changes in integration protocols. The
volatility of the tooling is itself a part of the moderni-
sation problem and the toolchain itself should be con-
sidered part of the environment to secure and evolve.
The use of modelling has opened new doors, es-
pecially to go beyond the mere architectural refactor-
ing by reconnecting from the solution into the prob-
lem domain, enabling the modernisation process to
recover and re-engineer the requirements prior to the
work at the architecture level. This relies on an elic-
itation process sharing some similarities with a reim-
plementation process. Although this work can gener-
ate overhead in the first iteration, it can be kept under
control by making sure the scope stays closed to the
system being modernised (avoid modelling connected
systems beyond their interactions). They should also
be kept at the level of abstraction of the considered
systems and leave more technical requirements for the
architectural phase. As the process is iterative, the re-
quirements of specific subsystem can be investigated
in due time. Finally, provided the model is main-
tained, each iteration with rely on model of increasing
quality and require decreasing effort to build, provid-
ing more incentive to fully exploit it. Note also that all
iterations may not need deep reengineering and stay
at a more technical level. In this case the ”horseshoe”
curve depicted in Figure 2 is simply lower and require
less steps in the process to achieve.
5 CONCLUSION & NEXT STEPS
In this paper, we proposed a methodology for driv-
ing the modernisation of software intensive systems
in an incremental way by progressively bridging the
gap between the as-is and to-be situations through
a model-based (MBSE) approach. We validated the
feasibility of the approach on a proof-of-concept case
focusing on a specific functionality of a railway com-
ponent. The rich set of models allowed us to go be-
yond pure technical and architectural considerations
and reason on the gap at problem level. On the other
side, it was also possible to extend the use of mod-
els down to the implementation using code generation
and efficiently requalify the modernised component
through automated testing with few issues to fix.
Although encouraging, this validation is still lim-
ited in scope and a number of improvements have
been identified as future work. The increment gran-
ularity needs to be assessed based on the cost vs time
to issue a milestone vs the global modernisation time.
A number of internal and external dependencies may
also need to be integrated in the project, e.g. the deci-
sion to migrate to a new technology or language. On
the tooling side, the need of reliable connectors across
the toolchain and of a global versioning system was
identified.
REFERENCES
Comella-Dorda et al. (2000). A survey of black-box mod-
ernization approaches for information systems. In
Proc. Int. Conf. on Software Maintenance.
Holladay, J. B. et al. (2019). Mbse infusion and modern-
ization initiative (miami): ”hot” benefits for real nasa
applications. In 2019 IEEE Aerospace Conference.
Johnson, J. and Mulder, H. (2021). Endless moderniza-
tion: How infinite flow keeps software fresh. Standish
Group.
Khadka, R. et al. (2015). Does software modernization de-
liver what it aimed for? a post modernization analysis
of five software modernization case studies. In 2015
IEEE International Conference on Software Mainte-
nance and Evolution (ICSME), pages 477–486.
Mathworks (2019). Simulink v10.
www.mathworks.com/products/simulink.html.
Morkevicius, A., Bisikirskiene, L., and Jankevicius, N.
(2016). We Choose MBSE: What’s Next? Complex
Systems Design & Management.
OMG (2005). System modeling language. http://www.omg.
org/spec/SysML.
OMG (2016). Architecture driven modernization. https:
//www.omg.org/adm.
PTC (2020). Windchill modeller v12.
www.ptc.com/fr/products/windchill/modeler.
Respect-IT (2002). Objectiver- goal-oriented requirements
engineering toolset. www.objectiver.com.
Santos, B. M. et al. (2018). Software refactoring for system
modernization. IEEE Software, 35.
Stavru, S., Krasteva, I., and Ilieva, S. (2013). Challenges
of model-driven modernization: An agile perspective.
Proc. of the 1st Int. Conf. on Model-Driven Engineer-
ing and Software Development.
van Lamsweerde, A. (2009). Requirements Engineering -
From System Goals to UML Models to Software Spec-
ifications. Wiley.
Weilkiens, T., Scheithauer, A., Di Maio, M., and Klusmann,
N. (2016). Evaluating and comparing mbse method-
ologies for practitioners. In IEEE International Sym-
posium on Systems Engineering (ISSE).
William M. Ulrich, P. N. (2010). Information Systems
Transformation: Architecture-Driven Modernization
Case Studies. Morgan Kaufmann.
Towards Incremental Model-Driven Software Modernisation: Feedback from an Industrial Proof of Concept in Railways
659