During the CAMPaM week, McGill colleague Jörg Kienzle organizes his Aspect-Oriented Modeling workshop at Bellairs, so there will be plenty of opportunities to interact with some very smart AOM people, which opens the possibility for further collaborations and cross-fertilization.

Computer Automated Multi-Paradigm Modelling (CAMPaM)

CAMPaM acknowledges that modelling is the central activity in and main enabler for the analysis and design of complex systems. Because of the heterogeneous nature of for example embedded systems for example and the many implementation technologies, multi-paradigm modelling is a critical enabler for holistic design approaches (such as mechatronics), to avoid overdesign and to support system integration. Multi-paradigm techniques have been successfully applied in the field of software architectures, control system design, model integrated computing, and tool interoperability. Eight CAMPaM workshops at Bellairs, many conference sessions and MoDELS '06, '07, '09, '10, '11 workshops have been held and a MoDELS 2012 MPM workshop is planned. A special issue of the journal Simulation was recently devoted to CAMPaM. See the (to-be-updated) CAMPaM page for more related material.
Multi-Paradigm Modelling spans the study of physical as well as software systems and combinations thereof. It adresses and integrates two orthogonal research dimensions:

1. model abstraction, concerned with the (refinenment, generalization, ...) relationships between models at different levels of abstraction;
   
   
2. multi-formalism modelling, concerned with the coupling of and transformation between models described in different formalisms.
   
   

To support the above, the following enabling theories/methods/technologies are considered crucial:

1. Modelling language engineering and in particular meta-modelling, concerned with the description (models of models) of classes of models. More explictly, the specification of formalisms (including their semantics -- note that language engineers usually reserve the term meta-model to a model of abstract syntax of a formalism). Taking meta-modelling one step further, the structure, look, and behaviour of complete formalism-specific modelling environments is specified and the environments are automatically synthesized.
   
   
2. the explicit modelling of transformations, treating transformations as first-class models. This leads quite naturally to questions about (meta-)model evolution, higher-order transformations (transforming transformations), co-evolution of models, multi-view modelling and syntactic and semantic model consistency.

CAMPaM explores the possible combinations of the above notions. It combines, transforms and relates formalisms, generates maximally constrained domain- and problem-specific formalisms, methods, and (visual) tools, and verifies consistency between multiple views.

1. The diversity in the research subjects of the attendees provides a fertile ground for cross-correlating research. In particular, since 2008, several of the workshop participants are not Computer Science researchers, but rather domain-experts (mechanical engineering, embedded systems, ...). The result of this interaction will be the application of methods and techniques that are well-known and established in different fields of research (such as meta-modelling, graph transformation, domain-specific modelling, visual modelling environments and component-based modelling) and will lead to cross-disciplinary collaboration. Furthermore, it should make evident the need for advances of research along avenues otherwise overlooked.
   
   
2. A concerted effort of the attendees will result in a consolidation of scattered CAMPaM-related work as well as a common vision on how to best evolve the field of CAMPaM. This vision will include detailed technical perspectives, joint publications, how Multi-Paradigm Modelling and Model-Driven Engineering may be introduced in education, as well as organizational plans.
   
   

We plan to focus on (some of) the following subjects during the workshop:

1. Abstraction and how it relates to concepts such as approximation and realization. In particular, how abstraction by means of "wrapping" legacy components is a powerful enabler for domain-specific modelling.
2. Foundations of domain-specific modelling with a particular focus on visual modelling and the synthesis of complex user interfaces.
3. Complex applications and how they on the one hand elicit new CAMPaM challenges driving new research and on the other hand apply and test/validate current CAMPaM state-of-the-art theory, techniques and tools. Application domains of particular interest this year are mechatronics and System on Chip design. Other topics may come up depending on the interests of the participants.
4. Foundations of CAMPaM/DSM:
   Meta-modelling and model transformation as enablers for multi-formalism and multi-abstraction modelling. The critical subjects (for the succesful large-scale adoption of CAMPaM) are (meta-)model evolution, (automated) model/transformation testing, model exchange, model debugging, multi-view modelling, model consistency, as well as design-space and deployment-space exploration.
5. Foundations of multi-formalism modelling and simulation. In particular, transformation and execution frameworks for models using multiple Models of Computation (MoCs).

In contrast to previous workshops, we will mostly, in addition to a few general presentations, work in small groups (as few as 2 participants) on specific problems. The results will be discussed globally during the evening sessions (7-10pm). Such focused discussion are likely to lead more directly to joint publications.