Background

I graduated from the University Paul Sabatier III in Toulouse with an MSc in Software Engineering" specialized in the methodology and techniques for modeling and validating embedded systems. After three years in the IRISA (Institute of Research in Computer Science and Random Systems) lab as a PhD student under the supervision of Jean-Marc Jézéquel I defended my thesis on the topic of Model and Metamodel Composition " entitled Separation of Mapping and Interpretation for Unifying Existing Model Composition Techniques.

In January" 2012 I joined the ESEO Group for a 6-month postdoctoral position to mature a research prototype with Olivier Beaudoux. In September 2012 I was hired as an associate professor in ESEO " Angers.

Research

Within the global framework of Model-Driven Engineering" which promotes models as first class entities I am interested in the implementation of executable models. Shifting practices from models as documentation to models for production executable models are actual software engineering artifacts with precise operational semantics. These semantics can feed an interpretation or simulation engine to execute models and expect concrete results. The scope of executing models encompasses system synthesis system verification and validation " or modeling at runtime.

Modeling complex system needs the two core principles of modularity and Separation of Concerns to understand the original problem and to decomposition the problem into manageable issues. These principles have a  strong impact on the modeling activity since they promotes the participation of multiple models and potentially multiple languages in the design of the global system. Flexible and customizable collaboration of models is thus required to adequately and efficiently answer to the original problem.

My research was therefore focused on the design of techniques for composing models and metamodels (i.e. domain-specific languages). Within the multitude of model composition approaches published in the literature" " I extracted the key concepts of model composition to propose a new definition of model composition that should improve reuse and flexibility in the design of new model composition frameworks.

Interests

From the contribution of my thesis" " I am interested in going further towards the formalization of a general definition of model composition as well as evaluating its impact on the design of model composition frameworks. This research spans from the observation of existing model composition approaches to the identification of behavioral units for the description of the operational semantics of model composition languages. My strong belief is that we should be able to build entire model composition frameworks by composing behavioral units of operational semantics that collaborate to achieve a given purpose.

To achieve such exciting breakthrough" " we should observe Psychology and Social Sciences domains to identify interesting collaborations with competent researchers from these fields.

Projects

I participated in the MOPCOM-I project from 2008 to 2011 as partner for the coordination of the WP2.1(models verification) and WP4 (reliable fusion of metamodels). We collaborate particularly with Thomson/Technicolor on: (1) the integration of heterogeneous protocols for the management of digital video broadcasting equipments; (2) the design of a tool-chain for the verification of the adaptation mechanisms synthesized for the integration.

We are currently working towards integrating the GEMOC initiative as principal partner to help in the definition of frameworks for the collaboration and integration  of heterogeneous software languages.

Prototypes

Kompose

We have made experiments with Kompose" which is basically a model composition tools for merging homogeneous models. Experiments led to some improvements in the tool particularly to support the mathematical union and intersection of homogeneous models. Last versions have been released on the Kermeta website. If you need additional information or expertise on Kompose " please contact me.

ModMap

The ModMap framework emerged from the novel definition of model composition proposed in my thesis. In collaboration with Thomson/Technicolor and as a proof-of-concept for the validation of the methodology" we implemented a prototype. Based on the theoretical concepts of the thesis " we derived a model composition framework which targets the integration of heterogeneous languages. The ModMap prototype is composed of code generators and graphical editor for the alignment of heterogeneous models. Current version of ModMap supports:

    • the extraction of ECore models from Java source code
    • the design of mappings between ECore model elements (GMF graphical editor)
    • the generation of a Kermeta model from the model of mappings that implements a integration strategy relying on a set of adapters.
    • the compilation of the integration model (i.e. Kermeta model) into Scala for execution
Versions of ModMap can be dowloaded here and future released will be proposed on this very page.

LOA

The LOA framework was originally designed by Olivier Beaudoux to support the specification and synthesis of rich graphical components and user interfaces. As part of my postdoctoral position" I worked towards the maturation of the original prototype " providing also a graphical editor for the specifications of UI.

We expect to provide a public release for the LOA framework very soon. If you are interested in the approach" " please contact Olivier Beaudoux or me.


Publications

Please refer to my Google Scholar profile for a curated list of published papers.


Courses

As associate professor in ESEO" " I currently participating in the following courses:

  • Introduction to Design Patterns
  • Advanced concepts in Java Technology
  • Foundations in Computer Science
  • Object-Oriented Modeling and Programming
  • Data and Systems Modeling