Multi-paradigm modelling is at the heart of MSDL research. It combines:

• Multi-formalism modelling
• Multi-abstraction modelling
• Meta-modelling (of both syntax and semantics)

Have a look at the Computer Automated Multi-Paradigm Modeling page, at the twelve Bellairs CAMPaM workshops '04, '05, '06, '07, '08, '09, '10, '11, '12, '13, '14 '15, and at the MoDELS '06, '07, '09, '10, '11, '12, '13, '14 MPM workshops.

• Official COST website of Action IC1404 "MPM4CPS"
• mpm4cps.eu project website

• Official COST website of Action IC1404 "cHiPSet"

• ITEA project fiche

• NECSIS project website
• project announcement

AToM³ is a CaMpaM tool the MSDL is currently developing. Have a look at AToM³'s preliminary home page.

Note that AToM³ is currently being re-designed from the bottom up. At the basis are hierarchical graphs (see Marc Provost's page for a description of the Python Graph Kernel PyGK), the system is independent of any GUI, everything is truly a model (including graph grammars) and can thus be transformed, and the system is mostly bootstrapped.

[twiki] Belgium is one of the first countries in the world to have started rolling out a national electronic ID card. In the current design, each e-ID card stores a single public key certificate that is tied to the national number and that is used in all transactions with government services. This approach brings serious privacy risks for citizens. The ADAPID project aims to design secure e-ID card applications that protect the privacy of citizens; some of these designs will function on top of the current e-ID card, while others requires design improvements in the e-ID card architecture itself. The adapID research activities span the following areas: digital credentials (how to prove that one is entitled to a service or without revealing one's identity), secure integration of biometry and cryptology, reliable dispute handling, trusted modules for securing applications, and services and legal aspects of trust in an open network. The new solutions will be validated for three key applications: e-archiving, e-health and e-government. For each of these a proof of concept will be implemented. McGill will contribute to the fundamental research (with a focus on model driven approaches) and will play a major role in the work for the e-health application.

The transformation between an Ordinary Differential Equation (ODE) model (possibly specified in the form of a Causal Block Diagram) and a behaviourally equivalent DEVS discrete event model has several merits.

• Efficient (optimistic) parallel simulation can be done using standard discrete event simulation techniques (such as Time Warp). Direct parallellisation from the ODEs is far less straightforward.
• Hybrid modelling where discrete event and continuous models are combined poses many challenges both conceptually and from the point of view of efficient simulation. Mapping ODEs onto DEVS reduces a hybrid model to an exclusively discrete event one.
• In the transformation from ODEs to DEVS, the state space rather than time base is discretized which has interesting implications in for example computer game applications.

For a short description of our work, see Jean-Sebastien Bolduc's doctoral proposal (pdf).

Recently, this work has evolved into not only explicitly modelling and simulating systems under study, but also different types of experiments.

In this project we investigate a general framework to support multi-paradigm modelling of systems that have an evolving structure. The aim is to develop a uniform method (and a tool) for adapting formalisms to deal with dynamic structure, rather than manually designing their variable structure variants. The focus is on DEVS and Statecharts (and variations and combinations thereof). For more information visit Ernesto Posse's website.

In July 2001 Profs Karel Driesen and Hans Vangheluwe were awarded over 100,000$ (matched in equal amounts by McGill University and the Province of Quebec adding up to a total of over 300,000$) equipment by the Canada Foundation for Innovation (CFI) for the establishment of a Modelling, Simulation, and Adaptive Computation infrastructure (mainly a Beowulf cluster with 32 dual-core nodes and 8Gb central memory connected by a Gigabit network switch). A short description of the project (and pictures) will be put up soon.

Extend (to make it a multi-formalism and meta-modelling language) the semantics of Modelica, a de facto standard modelling language for physical systems modelling. In his M.Sc. thesis, Steven Xu has built a research prototype Modelica compiler.

WEST++ (core, designed by Hans Vangheluwe and the WEST++ team at BIOMATH) and its commercial implementation Hemmis WEST form the basis for the next generation of Waste Water Treatment Plant modelling. It is an excellent example of a Domain-Specific Modelling and Simulation Environment. As such, it acts as a example for our recent research into the use of meta-modelling and model transformation for the generation of Domain-Specific environments. We plan to use meta-modelling and model transformation to generate a much improved WEST++.

Currently focused on the automatic generation of GUIs from Statechart specifications (modelling in AToM³, simulation in SVM and code generation using SCC). This will allow bootstrapping of the AToM³ GUI. There is extensive documentation on SVM and SCC. For those interested in more in-depth information on SVM and SCC, have a look at Thomas Feng's M.Sc. thesis.

A pertinent example of the use of Statecharts is in Denis Dubé's M.Sc thesis (Chapter 3). It describes how hierarchies of Statecharts can be used to synthesize visual modelling environments, including layout.

© Francescu Quilici

MSDL's contribution to GRaM is currently in synthesis of (1) efficient AI code from DCharts and (2) efficient physics simulation code from Modelica.

• MITACS internship programme
• NSERC Collaborative Research and Development Grants (CRD)
• NSERC Strategic Project Grants