Model-driven engineering, dynamic adaptation,
Numerous real time and/or embedded computer applications necessitate to combine software components to produce an executable processing chain. This raises two related concerns. First, the chain has to be configured before running. Second it has to adapt to changing environment by tuning components or modifying their assembly. A currrent trend in Software Engineering is to rely on models for both concerns. This model-driven approach is now accepted for the configuration part. A more speculative line of research targets the use of Models at Run Time (MRT) for the execution part.
For many years we have developed a knowledge-based approach, named Program Supervision (PS), to monitor and control complex program executions. This led to fully operational systems based on common frameworks, written in C++, together with domain specific languages and tools. Application domains were as different as image understanding for astronomy, traffic control, biology, or hydraulics.
This work has much in common with Models at Run Time, but because of its pionneering nature, it could not benefit from the up to date techniques in software modelling. The purpose of the internship is to investigate the mutual benefits of both approaches (PS and MRT) and to merge them into a modern technique taking into account modeling as well as reasoning aspects, from configuration to run time. The intend is of course to take advantage of the existing frameworks and tools, adapting them to the model-driven context. We are already working on a test application in Video Surveillance, a demanding domain in terms of dynamic adaptation.
|Work to be
The student will have to achieve the following tasks:
Good knowledge of C++, software modeling (UML-like). Model-driven would be a plus. No experience required in image or video processing.
PULSAR Team, INRIA-Sophia Antipolis.
EPI PULSAR, INRIA-Sophia Antipolis
2004 route des Lucioles BP 93
06902 Sophia Antipolis Cedex
Email : Sabine.Moisan@sophia.inria.fr
Tel: 04 92 38 78 47