AI models can now continually learn from new data on intelligent edge devices

A novel approach enables on-device training of machine-learning models on edge devices with limited memory, such as microcontrollers. This might enable edge devices to continuously learn from fresh data, removing data privacy concerns while allowing for user personalization. Microcontrollers, which are small computers that can execute basic commands, are the foundation for billions of linked devices, ranging from internet-of-things (IoT) devices to automotive sensors. However, because low-power microcontrollers have relatively little memory and no operating system, training artificial intelligence models on “edge devices” that operate independently of central computer resources is difficult. Read More…