Scholarlink

This paper introduces a novel health monitoring system designed to be worn on the wrist, which merges continuous physiological sensing with advanced personalization algorithms. The device employs photoplethysmography to monitor heart rate and oxygen saturation, incorporates digital temperature sensing, and utilizes inertial motion tracking to deliver extensive health insights. In contrast to current consumer wearables that depend on static thresholds and cloud-based processing, our system features on-device baseline learning algorithms that adjust to the unique patterns of individual users. The accompanying Android application includes an adaptive reminder scheduling system that learns user behaviour patterns to enhance notification timing and boost adherence rates. The hardware is constructed from commonly available components, such as the ESP32 microcontroller, MAX30102 optical sensor, DS18B20 temperature sensor, and MPU6050 inertial measurement unit. All data processing and personalization take place locally, safeguarding user privacy while facilitating offline functionality. Experimental validation indicates heart rate accuracy within 2.1 bpm, step counting accuracy surpassing 97%, and sleep detection accuracy of 87% when compared to manual sleep logs. The adaptive reminder system demonstrated a 23% increase in user adherence compared to fixed-time notifications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Personalized Health Monitoring Wearable with Intelligent Baseline Learning and Adaptive Reminder System

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • R. Banu Prakash,
  • V. Dhilipkumar,
  • Isha Raut,
  • Yatnesh Shankar Rampelli,
  • Oana Geman,
  • Roxana Toderean,
  • Octavian Postolache

摘要

This paper introduces a novel health monitoring system designed to be worn on the wrist, which merges continuous physiological sensing with advanced personalization algorithms. The device employs photoplethysmography to monitor heart rate and oxygen saturation, incorporates digital temperature sensing, and utilizes inertial motion tracking to deliver extensive health insights. In contrast to current consumer wearables that depend on static thresholds and cloud-based processing, our system features on-device baseline learning algorithms that adjust to the unique patterns of individual users. The accompanying Android application includes an adaptive reminder scheduling system that learns user behaviour patterns to enhance notification timing and boost adherence rates. The hardware is constructed from commonly available components, such as the ESP32 microcontroller, MAX30102 optical sensor, DS18B20 temperature sensor, and MPU6050 inertial measurement unit. All data processing and personalization take place locally, safeguarding user privacy while facilitating offline functionality. Experimental validation indicates heart rate accuracy within 2.1 bpm, step counting accuracy surpassing 97%, and sleep detection accuracy of 87% when compared to manual sleep logs. The adaptive reminder system demonstrated a 23% increase in user adherence compared to fixed-time notifications.