Increasing the space-time precision of the operational picture for strategic superiority is a major asset for the safety of our troops. Innovative furtive electronic agents have to sense and collect the right information without overwhelming analysts and communication media.
Left on the battlefield, Unattended Ground Sensors (UGS) have to provide long-term autonomous sensing of their environment. Bio-inspired multisensor systems are the best compromise between energy consumption and sensing efficiency (B-SAVED).
In order to quickly anticipate threats with UGS, innovative algorithms must be optimised to fit with embedded hardware computational resources (SmartCam).
Understanding the situation at the sensor level requires local knowledge, real-time and accurate event assessment by maximising the number of recognitions per second & per watt (e.g. TEAM project).
Facing highly evolving situations implies self-adaptive UGS. It is a major goal to design evolving hardware through autonomous learning.
- Development of autonomous sensors for zone surveillance
- Bio-inspired multiple-sensor systems
- Algorithms for the autonomous learning of sensing systems
B-SAVED concept
Bio-inspired Smart Autonomous Visual Event Detector
SmartCam
Autonomous virtual expert detecting visual events and alerting when necessary
I-WARN: Intelligent Watchdog with Autonomous Recognition and Networkability