PhD StudentsAurélien Massein 2015-2018
Design of Sensor Network for Human Activity Monitoring in Indoor Environmentsstrong> We investigated on evaluating sensor network quality for indoor activity monitoring. We based our criterion on intermediary informations related to the person's actions within its indoor environment, such as its presence in precise regions or its interaction with specific objects. More precisely, we narrowed our problem on studying reliability of presence mainly and interaction features of an action using the notion of region from set arithmetic. We therefore defined two close-linked notions : Interest Region for relevant Human positions or uses for an action and Coverage Region effective space measured by sensors. We combined those two notions into a Confusion Matrix, and applied the F-measure to express reliability of an action's feature. Network's quality is maximal when a relevant feature is well-covered and well-detected by sensors. As we have several reliabilities, we resorted to generate a set of pareto non-dominated solutions.
Current conferences organisation
FP7 - ICT - 2013.5.3 RAPP
[2013-2017] An open-source software platform for developers to easily create and deliver robotics apps dedicated to social inclusion of elderly people.
The ULLO partner startup (http://www.ullo-world.fr) is working on connected textile for medical applications. However, within PosTex, it is necessary to adapt its composition to the desired properties and master the techniques of assembly (sewing) to maximize the identifiability of human movements and reduce embedded electronics.
AIO is the key player for European assembly industries in their various CSR approaches. This “smart low tech” company excels in the foremost European automobile constructors and the massive aeronautical companies, who in turn are opting for more efficient, more human, more “health and safety” and greener factories of the future.