Continuous monitoring of vital signs during daily life gives potential for early diagnostic and will provide guidance for healthier lives.

Our time series analysis experts leverage the latest advances in deep learning to extract elusive information from optical blood flow sensors.

7 Sensing Software unleashes the potential of PPG sensors and embedded processing optimizations to deliver early detection of minute variations in vital signs.

TinyML and AutoML tools are not designed for this high complexity use case (Deep Neural Network with tens of thousands parameters).

Therefore we extended the TinyML concept to support complex Deep Learning Models on Microcontrollers : our solution is optimized for Cortex-M4 @ 64MHz for ultra-low footprint and compute load: 82,3 kB Flash and 44,5 kB RAM / 7,56% of CPU load.