STMicroelectronics — ST1VAFE3BX Bio-Sensing Chip

STMicroelectronics has introduced a new biosensing chip which performs motion sensing and physiological signal interfacing at very low power in an ultra-compact form factor.  

Thanks to its size, integration and power consumption, the ST1VAFE3BX provides new options for the design of next-generation health-monitoring products in wearable form factors such as smart watches, sports bands, connected rings, and smart glasses. Measurement functions supported by the chip include heart-rate variability, cognitive function, and mental state.  

The ST1VAFE3BX is supplied in a 12-lead LGA package which has a footprint of 2 mm x 2 mm.  

The ST1VAFE3BX provides a complete vertical analog front end (vAFE) for biopotential sensors: this simplifies the detection of vital signs which can indicate physical or emotional state. The vAFE can support biosensing functions such as electrocardiography (ECG), electroencephalography (EEG), seismocardiography (SCG), and electroneurography (ENG), the measurements of which can be used to reliably indicate health status or physiological responses to events such as stress or excitement.  

As well as providing a precision biosensing front end, the ST1VAFE3BX also includes an accelerometer for inertial sensing of the wearer’s movement. The output from the accelerometer can be synchronized with the biopotential sensing signals to help the application infer any link between measurements and physical activity. 

The ST1VAFE3BX also integrates the ST machine-learning core (MLC) and finite state machine, which enable product designers to implement simple decision trees for neural processing on-chip. These AI capabilities let the sensor handle functions such as activity detection autonomously, lifting this processing burden from the host CPU, accelerating system responses, and minimizing power consumption.  

Development with the ST1VAFE3BX is supported by ST software tools such as MEMS Studio in the ST Edge AI Suite, which engineers can use to configure decision trees in the MLC.