My thesis is entitled “Contextual integration of heterogeneous data in an open and opportunistic smart environment: application to humanoid robots” or “Intégration contextuelle de données hétérogènes dans un environnement ambiant ouvert et opportuniste : application aux robots humanoïdes” in Molière’s language (’cause yes, i’m french… cocorico !). To shorten this magnificent title: i’m working on robots in smart homes in a general way. I tackle several issues implied by the interaction of a robot and a smart environnement and rely on semantic technologies, in particular ontology.
To do so, i defined a very simple and non innovative robot architecture to structure my thesis:
- Perception: The robot perceives its surrounding and uses the available devices to have the most complete knowledge. Challenges in this layers are the uncertainty of sensors and the open (sensors discovery, integration and management of disconnection) aspect of the environment.
- Cognition: The robot understand what’s happening. The idea is to use situation awareness techniques. Handling uncertainty, of both data and reasoning is a problem. Learning new situations and/or reacting to non given one is also a great challenge.
- Action: The robot acts to reach a goal and overcome a situation. To do so, it (he ? she ?) uses a task planner. Dynamism, opportunism and context acquisition (linked to perception) are challenges in this layer.
- Bonus: And of course, connecting these three layers can offer possibilities and challenges. For instance, using situations to adapt the task planning phase is an interesting way to explore. In the end, building the whole working architecture is the target goal of my work.
You can find more about my work in the others pages. Don’t hesitate to contact me for further information: nathan.ramoly@telecom-sudparis.eu
I’m awesomely supervised by Amel Bouzeghoub and Béatrice Finance.