Smail joined IRT SystemX in November 2013 to begin a thesis on the subject of multiple-objectives optimization by composition of model transformations. He talks about his research in the FSF (Safe and Reliable Embedded Systems) project, under the tutelage of Télécom ParisTech.
What was your thesis subject?
During my thesis, I examined the field of multi-objective optimization and non-operational properties such as reliability or response time in a railway system. A railway system includes several material components such as control units and software components (operations executed by the control units, such as door opening on a train); and a whole set of non-operational properties. I tried to find an optimum model by grouping the various hard and software to improve all the non-operational properties.
The problem I discovered was that improving one property may damage one or several other non-operational properties. It is difficult to find a model that optimizes all the non-operational properties simultaneously.
So for my thesis I made a framework enabling all the non-operational properties in a railway system to be optimized the best possible. I applied genetic algorithms and model transformation to do this. By combining them, I was able to provide just five or six solutions, out of millions of possibilities or models, that best addressed the compromise between various conflicting non-operational properties.
What do you remember about your work?
On the human side, I met very interesting people in the FSF project and the other projects at the IRT. I met many PhD students and research engineers, because they were open to helping other people.
On the scientific side, I particularly enjoyed the Heads Up meetings organized almost every week and which helped us discover various fields.
IRT SystemX also brought me many contacts with the industrial partners of the FSF project. They all helped me throughout my thesis.
What is your best memory at SystemX ?
We celebrated the first anniversary of the FSF project in 2015, along with all the team. It was a really enjoyable event.
What happens next?
My PhD opened the world of research for me. I hope to move into industry in the field of validation and development of railways systems.
More about Smail Rahmoun
Thesis subject: Multiple-objectives optimization by composition of model transformations
R&D project: FSF – Safe and Reliable Embedded System
Qualification: Master 2 “Shared systems and applications”, Université Pierre et Marie Curie.