This research is focused on microsystems integration in new systems with the challenge of implementing a smart atmosphere based on the breakdown/distribution of microsystems with autonomous communication, energy and decision. Our research efforts are focused on the design of nodes, or on the integration of systems built around MEMS that are either developed internally (see Advanced MEMS) or used as components on the shelves. In the latter case, it is necessary to prove and qualify the reliability of the microsystem in real conditions. Hence, we contribute to the development of a reliability physics, and we have been involved in the European projects PATENT on "MEMS and RELIABILITY”, in POLYNOE, and we patented our developments with CNES.

Our system design methodology is based on a virtual prototyping approach in which we use multi-scale models developed in the group. We ultimately propose VHDL and/or VHDL AMS multi physics system models, which take into account the MEMS in its environment and its operating conditions, and we aim at predicting the reliability of the system before any manufacturing phase. Note that the CAD platform is extensively used, and we share our knowledge through VHDL and COMSOL models.

To clarify the scope of our field investigation on Wireless Sensor Networks, we have not been developing transceivers, we rather use a communication channel, and we integrate the algorithms of multi sensory fusion in order to diagnose abnormal function. Finally, the issue of energy recovery is harvested based on mechanical and/or acoustic vibration, and we are working with specialists in storage and conversion. The embedded system applications, driven by the proximity of AESE and CancerBioSanté “Pôle d’excellence”, are focused on industrial partnerships on the observation and diagnosis of health: 

• Sensor level: we have developed a new generation of presence detector based on bolometry for direct body temperature measurements. This approach is relevant for old people with slow displacements. It has been industrially protected with EDF,
Figure 8: (a) Matrix thermopile sensor 2 r 8 (b) Micro-bonded sensor on support
• Diagnosis level: we propose autoadaptative experimental models representing the users habits. The more complex algorithm has been developed for the PROSAFE project in the case of patients with Alzheimer disease.
• Experimentation level: we have implanted a complete system with a network connecting residents and doctors, and validated it with Toulouse-Muret Hospital, isolated housing (Orléans), and retirement pensions (Tibiran-Jaunac).

Fig. 8 : (a) Matrix thermopile sensor 2x8 (b) Micro-bonded sensor on support

• We develop a new WSN based on motion sensors, and on accelerometer signal processing. This technology has been transferred to a new company: TAG TECHNOLOGIES (created in 2006, 10 persons working today, 5 PhD students from LAAS).

Figure 9: (a) Node motion sensor (b)Power harvesting module design.

• we develop a mechanical piezo harvesting node dedicated to the control of aeronautical structures (AIRBUS FRAE Autosens Project).