A power efficient wide operating range accelerometer with in-sensor compute
Matteo Fusi, Product Marketing Director, STMicroelectronics
Understanding Acceleration
Acceleration is a fundamental physical quantity. For an object in free-fall, certain quantities are equal to zero. For example, the net force acting on the object is zero, assuming air resistance is negligible. This leads to a constant velocity, not zero acceleration.
Orders of Magnitude (Acceleration):
- 2000 Toyota Sienna: 0 to 100 km/h in 9.2 s (approx. 0.3075-0.314 g)
- Standing on Earth at sea level: standard 1 g
- Maximum for human on a rocket sled: 46.2 g
- Formula One 1994 Monaco Grand Prix Karl Wendlinger Crash: Not specified, but typically very high
Acceleration levels can range significantly:
- 10g - 50g
- 50g - 170g
- 50g - >200g
- >45g
- >50g
Introducing the LSM6DSV320X
The LSM6DSV320X is a new sensor for accurate and efficient measurement of both high and low acceleration levels. It features embedded intelligence for power-efficient management of all sensing channels.
Key Features:
- Edge AI and self-configuration:
- Finite state machine (FSM)
- Machine learning core (MLC)
- Adaptive self-configuration (ASC)
- Sensor fusion low power (SFLP)
This IMU (Inertial Measurement Unit) is designed with a dual MEMS accelerometer, offering industry-first high performance for applications requiring intensity, impact, and shock detection.
The multi-core architecture enables advanced partitioning of data processing:
- Primary output: UI (User Interface), EIS (Electronic Image Stabilization), High-g (for A data)
- Auxiliary output: OIS (Optical Image Stabilization) (for A and G data)
The LSM6DSV320X supports primary output via SPI, I2C, and I3C, and auxiliary output via SPI and I3C.
Library for Fusing High-g and Low-g Data: MotionXLF
The MotionXLF library is designed for fusing high-g and low-g data. It manages the activation and deactivation of the high-g accelerometer for optimal power efficiency.
Data Characteristics:
- High-g data: Characterized by poor resolution and high dynamic range.
- Low-g data: Characterized by high resolution and poor dynamic range.
The system handles time-varying offset evolution, discontinuity, and noise profile inconsistency.
The Finite State Machine (FSM) plays a crucial role in managing the sensor's power states.
Acceleration Data Visualization:
A graph shows acceleration levels, with a high-g accelerometer measuring up to 320g and a low-g accelerometer measuring up to 16g.
Connecting Users with Their Environments
The LSM6DSV320X provides enhanced contextual awareness through IMU embedding AI and advanced processing features. This enables a variety of applications:
- Emergency calls activation
- Asset tracking & damage sensing
- Intense movements tracking
- Concussion monitoring
Applications by Device Type:
- Smartphone & Foldable: Emergency call, Activity recognition, Step counting, Lid angle detection, Image stabilization
- Wearable: Emergency call, Intense sports tracking, Wrist tilt, Activity recognition, Step counting
- Smart tags & IoT: Shock and drop impact tracking, Context awareness, Motion tracking
- PC, laptop, and tablet: High impact recording, Lid angle detection, In and out of bag, Lap versus desk
- Gaming: High impact recording, Intense gestures, Motion tracking
A Product for You and All Around You
The LSM6DSV320X is an industry-first IMU with a dual MEMS accelerometer, accurately measuring up to 320g full-scale range. It enables new applications for personal electronics and IoT, including emergency calls, drop & fall tracking, and concussion monitoring.
This miniaturized AI-enabled sensor offers context-awareness, reducing system-level complexity and enhancing efficiency.
