Seminar (organized by the research team ICI)

Romain GROSSE & Freddy KAMDEMS

PhD students at Heudiasyc

Tuesday, January 5, 2016 at 14:00 in the GI041 room
Pesentation of Romain Grosse

Interaction au regard dans les systèmes optroniques portables

Abstract :

La jumelle Infrarouge (IR) multifonction de Sagem Défense sécurité est un produit dont la complexité d’utilisation augmente en même temps que le nombre de fonctionnalités disponibles, si bien que la méthode d’interaction actuelle (boutons) arrive à ses limites. Nous proposons d’intégrer un mécanisme de suivi du regard dans ces produits comme modalité d’interaction. L’état de l’art reprendra le fonctionnement et les utilisations classiques des systèmes de suivi du regard ainsi que leurs limites en tant que modalité d’interaction. En effet, la simple instanciation de commande en utilisant le regard n’est pas un problème trivial. L’interaction multimodale regard-bouton est étudiée pour tirer profit des avantages des différentes modalités.

Presentation of Freddy KAMDEMS

Federation of Systems of Systems (SoS) models

Abstract : The mastering of complex systems goes through the modeling/engineering of the real system before it is really built. The modeling of these complex systems implies several different Domain-Specific Practices (DSP) composed by people, processes, tools, modeling paradigms, modeling languages, models etc. They tackle different concerns, have different life cycles and evolutions. Finally DSPs are autonomous, heterogeneous, distributed and sometimes independent. These characteristics are exacerbated for SoS modeling. Clearly, regardless that modeling environment, there is a need to get a comprehensive understanding of the real system thanks to its different models ; developed here with a Model-Based System Engineering approach. Connecting such autonomous, heterogeneous and diverse models is an old and topical issue yet urgent in the modeling of complex systems. This issue becomes crucial and critical for SoS modeling since the environment properties/constraints must be well regarded/complied with. A motivating example should illustrate a challenge at the level of a real system. After a review of related works, we discuss different practices and what is lacking. We argue on the need of a federated approach for connecting models. We present our approach, which is grounded on comprehensive and formal foundations stemming from Category Theory. We also discuss implementation questions and related frameworks. One benefit expected from a federation is to run some analysis on connected models. Hence, based on the running example, we present possible analysis approaches and some challenges which arise. This will lead to discussions and way forward


FR SHIC 3272

Collegium UTC/CNRS