Smart glasses need to move and track your head accurately without using much battery. This sensor makes AR, AI, and sports glasses work better and smarter.

Smart glasses face challenges in accurate motion tracking, stable image and video capture, and intuitive controls, all while maintaining low power consumption. InvenSense SmartMotion for Smart Glasses from TDK solves these problems, enabling developers to build AI-driven, augmented, and extended reality applications that respond reliably to head movements and user posture. It allows smart glasses to track orientation, recognize posture, stabilize images and video, and provide intuitive controls, helping devices deliver seamless experiences.
At the core of the solution is the ICM-45685 sensor. It combines on-chip sensor fusion software with optimized memory, allowing multiple algorithms to run at the same time. Key capabilities include high-accuracy head orientation tracking, image stabilization, and activity monitoring. Building on the widely adopted ICM-45686, the ICM-45685 adds software features specifically for smart glasses. Additional features, such as donning/doffing detection and voice vibration detection (VVD), are expected later this year.
TDK also provides a full ecosystem of sensors for smart and AR glasses, including PositionSense 9-axis sensors for orientation and navigation, TMR sensors for hinge detection, low-power MEMS microphones for voice control, ultrasonic time-of-flight sensors for wear detection and gestures, and temperature sensors for body, ambient, or optical component monitoring. These sensors work together to enable next-generation smart eyewear.
The main applications for SmartMotion include AI smart glasses, augmented or extended reality glasses, and fitness or sports eyewear. Its main features and benefits are high-accuracy head orientation tracking using the on-chip gyro-assisted fusion (GAF) algorithm, optical and electronic image stabilization (OIS/EIS) for camera modules, activity tracking via machine learning with the InvenSense sensor inference framework (SIF), superb vibration rejection with BalancedGyro technology, and an ultra-low-power IMU for extended battery life.







