ST Launches Inertial Measurement Unit with Machine Learning Core
When performing AI pre-processing of MLC motion data
Reduces load with 1/1000th of the power consumption compared to an MCU


STMicroelectronics released its latest ISM330DHCX and LSM6DSRX 6-axis iNEMO™ inertial measurement units (IMUs) on the 17th.

Through this, ST announced that it is extending the benefits of its Machine-learning Core (MLC) technology for motion detection to industrial and high-end consumer applications.

MLC performs simple AI preprocessing of motion data using about 1/1000 of the power required by a typical MCU to handle the same task. Consequently, this reduces the load on the host MCU, thereby extending the battery life of the context awareness and motion detection device, delaying maintenance, and reducing its size and weight.

The LSM6DSRX and ISM330DHCX are suitable solutions for high-end consumer and industrial applications such as AR/VR, drone flight control, dead-reckoning navigation systems, disk antenna attitude control systems, fleet management, container tracking devices, and dynamic inclinometers for industrial vehicles.

The consumer-grade LSM6DSRX includes a 3-axis digital gyroscope and a 3-axis accelerometer with an extended full-scale angular-rate range of up to ±4000dps. The industrial-grade ISM330DHCX provides stability over a temperature range of -40°C to 105°C, along with a 10-year supply warranty.

The MLC of each device interacts with the built-in Finite-state machine (FSM) logic and can perform simple iterative algorithms, such as step, hit, and rotation calculations, with lower power consumption compared to implementations using an MCU. The FSM detects a preset number of events or transmits a signal to the main controller after a set time has elapsed.

Both devices are in mass production; the ISM330DHCX is available as a 14-lead plastic LGA (Land grid array) device, with prices starting at $4 for 1,000 units. The consumer-grade LSM6DSRX starts at $3.50.