Abstract:Microcontroller Units (MCUs) are ideal platforms for edge applications due to their low cost and energy consumption, and are widely used in various applications, including personalized machine learning tasks, where customized models can enhance the task adaptation. However, existing approaches for local on-device personalization mostly support simple ML architectures or require complex local pre-training/training, leading to high energy consumption and negating the low-energy advantage of MCUs. In this paper, we introduce $MicroT$, an efficient and low-energy MCU personalization approach. $MicroT$ includes a robust, general, but tiny feature extractor, developed through self-supervised knowledge distillation, which trains a task-specific head to enable independent on-device personalization with minimal energy and computational requirements. MicroT implements an MCU-optimized early-exit inference mechanism called stage-decision to further reduce energy costs. This mechanism allows for user-configurable exit criteria (stage-decision ratio) to adaptively balance energy cost with model performance. We evaluated MicroT using two models, three datasets, and two MCU boards. $MicroT$ outperforms traditional transfer learning (TTL) and two SOTA approaches by 2.12 - 11.60% across two models and three datasets. Targeting widely used energy-aware edge devices, MicroT's on-device training requires no additional complex operations, halving the energy cost compared to SOTA approaches by up to 2.28X while keeping SRAM usage below 1MB. During local inference, MicroT reduces energy cost by 14.17% compared to TTL across two boards and two datasets, highlighting its suitability for long-term use on energy-aware resource-constrained MCUs.

[Wearable Medical Devices' MCU Selection Analysis Based on the ARM Cortex-MO+ Architecture].

[MCU Selection Analysis in Portable Electronic Medical Instrument Design].

Enabling Efficient Wearables: An Analysis of Low-Power Microcontrollers for Biomedical Applications

Design of Low Power Consumption Controller Applied to Implantable Medical Device

A low-power remotely-programmable MCU for implantable medical devices

An ASIC Solution for Intelligent Electrodes and Active-Cable Used in a Wearable ECG Monitoring System.

Ultra-low-power Physical Activity Classifier for Wearables: From Generic MCUs to ASICs

Design and Implementation of 16-bit RISC MCU for Medical Electronics Applications

A Software Development Platform for Wearable Medical Applications

Low Power Design of Portable ECG Monitor Using C8051F040 Microcontroller

The research and implementation of intelligent power management system based on MCU

An Energy-Efficient Soc For Closed-Loop Medical Monitoring And Intervention

Gate Oxide and Implantation Process Co-Optimization for Low-Power MCU Applications

Low-Power Wireless Wearable ECG Monitoring Chestbelt Based on Ferroelectric Microprocessor

Low Power Design in SOC Design of Embedded MCU

Low-power SoC design and system implementation for medical applications

Implementation and performance comparison of MMC firmware on RISC-V and ARM-based MCUs

Harnessing FPGA Technology for Energy-Efficient Wearable Medical Devices

A Low Power MCU Design for Battery-Less TPMS SoC

0.45v and 18pA/MHz MCU SOC with advanced Adaptive Dynamic Voltage Control (ADVC)

Low-Energy On-Device Personalization for MCUs