Research areas of S4M
1- Electronic design of generic and adaptive measurement nodes
2- Research for "signatures" of behavioral changes by combining signals
3- Signal processing and positioning algorithms for distributed objects
4- Original reconfigurable architectures
5- Optimization of interfaces of connected devices human/material
Our algorithmic approach relies on learning the behavior of the system to be observed, its behavioral modeling (statistical behavior, stochasticity, time, frequency ...) and on the detection of drifts by comparing the behavior observed at time t with respect to a learned reference model (model which is adjusted continuously).
Many demonstrations of these prototypes integrating this approach have been carried out in close collaboration with industrials and healthcare institutions: