Cyrille Martin

Printer-friendly versionPDF version

Contact Information

Office G120
Laboratoire d'Informatique de Grenoble (UMR 5217)
Magma and IIHM Teams
Maison Jean Kuntzmann (G120)
110 avenue de la Chimie
38400 Saint-Martin-d'Hères, France.
Phone: +33 4 76 51 46 38
Web: Pages personnelles (fr)

Research Interests

My research focuses on the Human-Centred Planning, in a context of application related to the Ambient Intelligence, and more particularly for the Composition of Interactive Systems.

The field of the Human-Computer Interface brings new scientific challenges to the field of Artificial Intelligence. Indeed, initially used by robots, planning could support the human to achieve its goals.

The planners must be reconsidered to answer still open questions, in particular with regard to the usability of the plans by humans. For that, I introduce a new concept: the flexibility in the plans for the end-users. I developped a graph-based planner during my PhD: LGP.


Short Biography

  • PhD in Computer Science (Human Computer Interaction, Artificial Intelligence) at Grenoble University

In a context of Ambient Intelligence, some of the user's needs might not be anticipated, e.g. when the user is in an unforeseen situation. In this case, it is possible that there is no system that exactly meets their needs. By composing the available systems, the user could obtain a new system that satisfies their needs. In order to adapt the composition to the context, the composition must allow the user to make choices at runtime. This means that the composition should include control structures for the user: the composition is flexible.
In this thesis, I deal with the problem of the flexible composition by automated planning for which I propose a model. The sequence and the choice operators are defined and used to characterize flexible plans. Two other operators are then derived from the sequence and the choice operators: the interleaving and the iteration operators. I refer to this framework in order to define the flexibility produced by my planner, LambdaGraphPlan (LGP for short), which is based on the planning graph. The originality of LGP is to produce iterations. I show that LGP is very efficient on domains that allow the construction of iterative structures.

  • Master in Computer Science (Human Computer Interaction, Artificial Intelligence) at UJF University