The application domains of this thesis are Health Smart Homes, and the research is more precisely centered on the improvement of daily-living for cognitively impaired persons and theirs caregivers.The proposed system can observe the context of each person, personalize the environment and assist the tasks detected if they need to be. Every action of the system is as unobtrusive as possible and takes into consideration the presence of more than one person. To personalize and assist the daily-living activities of a lone person, we need to know his personal context. This context is the conjunction of the preferences and habits, the illness or impairment, the movements in the smart home and the state of the various sensor and electrical devices, and the current activities that are detected for one person. To be able to assist many persons simultaneously, we need to compute the overall conjunction of each and every person's context since every presence can influence the global context and every personal one. This complexity brings a lot of problems like the multiple person localization and identification, or the personalization and assistance of multiple persons in the same space with various activities. Those problems are even more interesting since, following an ethical choice to ensure inhabitant's privacy, this project avoid the use of some intrusive technologies.
| Identifer | oai:union.ndltd.org:usherbrooke.ca/oai:savoirs.usherbrooke.ca:11143/5168 |
| Date | January 2011 |
| Creators | Castebrunet, Matthieu |
| Contributors | Rialle, Vincent, Giroux, Sylvain |
| Publisher | Université de Sherbrooke |
| Source Sets | Université de Sherbrooke |
| Language | French |
| Detected Language | English |
| Type | Thèse |
| Rights | © Matthieu Castebrunet |
Page generated in 0.0021 seconds