Section: Application Domains

Living systems: ecological monitoring, modelling, estimation and identification of biological systems, human-computer interaction

Modelling, estimation or detection for living is difficult because such systems cannot be isolated from external influences. Using our numerical differentiation tools, together with modelling techniques, we want to study the following four applications:

• Biosensing: Unlike classical approaches deploying physical sensors, biological systems can be used as living sensors. The marine biology lab EPOC (CNRS, Bordeaux) has developed underwater sensors for bivalve molluscs (such as oysters) measuring and sending through RGPS the opening gap between the two valves. We want to use it for water quality monitoring by either identifying oyster's rythm I/O models or by using our differentiation tools. Spawning detection is also considered (ANR WaQMoS).

• Human-Computer Interaction: Reduction of the latency between the human input and the system visual response in HCI (ANR TurboTouch). To do that, a simple forecasting algorithm for latency compensation in indirect interaction using a mouse has to be developed based on differentiators.

• Smart bracelet: Design a dynamical model for the GSR and for the development of an online algorithm making the GSR signal independent of the user movements. Most resulting computations should be embedded into the bracelet. Collaboration with NEOTROPE (start-up developing a bracelet intended for strong human emotion detection).

• Microbial populations: Real-time control of synthetic microbial communities (Inria Project Lab, COSY, under evaluation).