Abstract image of a laptop surrounded by interconnected data visualizations, light orbs, and infographic icons, depicting concepts of online security, global networking, data transfer, and digital communication, with a predominantly blue and purple color scheme

SmartEdgeML: Run machine learning models on sensors

Ultra-low power execution of pre-trained machine learning (ML) algorithms directly on IMUs

Woman sitting on a couch listening to music with wireless over-ear headphones

Make smarter devices with on-sensor machine learning

TDK SmartEdgeML brings the power of machine learning (ML) algorithms to motion sensors. Enable next-generation features such as activity and gesture recognition to smartwatches, hearable, wearables, and health monitors.

TDK SmartEdgeML brings ML algorithms to sensors

Key capabilities of SmartEdgeML

Ultra-low power

Run decision tree ML models on < 30 µA for always on sensing without sacrificing battery life

Precise detection

Easy activity and gesture detection through motion data machine learning

Auto mode & easy development

Developer tools to get started quickly, configure options, train, view performance metrics

Enable new use cases

Train sensors to recognize human activities, motions, and gestures

Ensure data privacy

Leverage highly personal user data without it ever leaving the sensor

Improve battery life

Machine learning can execute on sensors while the rest of the devices is asleep

Ultra-low power

Run decision tree ML models on < 30 µA for always on sensing without sacrificing battery life

Precise detection

Easy activity and gesture detection through motion data machine learning

Auto mode & easy development

Developer tools to get started quickly, configure options, train, view performance metrics

Enable new use cases

Train sensors to recognize human activities, motions, and gestures

Ensure data privacy

Leverage highly personal user data without it ever leaving the sensor

Improve battery life

Machine learning can execute on sensors while the rest of the devices is asleep

Add device ML intelligence to an array of applications

SmartEdgeML enables personalized experiences for a range of devices

Close-up view of blue eye in glasses with futuristic holographic interface to display data

Smart glasses

Low power gesture, movement and direction awareness

Run sensing algorithms on sensor for extra long battery life
Detect a variety of gestures and automated controls, including taps, donning and doffing, and voice detection

Holographic display of health metrics above a tech-enabled wrist watch

Wearables & hearables

Deliver accurate gesture and activity monitoring to enable new use cases

Reliably recognize and differentiate head, hand, and body gestures despite individual differences
Better differentiate between different activities and exercises

Smart glasses

Low power gesture, movement and direction awareness

Run sensing algorithms on sensor for extra long battery life
Detect a variety of gestures and automated controls, including taps, donning and doffing, and voice detection

Wearables & hearables

Deliver accurate gesture and activity monitoring to enable new use cases

Reliably recognize and differentiate head, hand, and body gestures despite individual differences
Better differentiate between different activities and exercises

Machine Learning sensors

Explore popular sensors that can run ML models

T5848

Bottom Port I²S Digital Output Multi‑Mode Microphone

The T5848 is a low noise, high AOP I²S microphone with a 24‑bit PCM output for direct connection to a wide variety of SoCs. It also features Acoustic Activity Detect (AAD), a new ultra-low power edge processing feature consuming as low as 20 µA.

View Details