Home
學生控制台
Uedu Fit
註冊會員/登入
研究知情同意書
Uedu myGPTs
Uedu Jupyter
學習特質探索
Learning Portfolio
山巒地圖
我的證書
教師控制台
課程設定
EMI Toolkit
Assessment Toolkit
Interaction Toolkit
Wellness Toolkit
Discussion Toolkit
AI 知識庫
功能介紹
教學實踐研究
學術交流
教學工作坊
課程搜尋
FAQ常見問答
學生研究團隊招募
建立AI助教說明
教師研究社群
教學研究支援
Uedu Labs
隱私權政策
資料安全
研究倫理
Uptime 數據

myGPTs

--

Jupyters

0

AI 回覆桌面通知

AI 助教回覆完成時顯示桌面通知

聊天訊息通知

同學在討論區發送訊息時通知

聲音通知

每當有新通知時播放提示音

Festival of Learning 2026

From Wearables to AI Tutors: A Hands-on Tutorial on Physiological-Aware Learning Analytics

從穿戴裝置到 AI 助教:生理感知學習分析實作教學

Half Day (3.5 hours) 2026 Chia-Kai Chang, Kuei-Hao Li
Session 2 Hands-on + Lecture Preparing Materials

Physiological Sensing & Paired Data Analysis

Garmin & UeduPAD HRV Collection, Uedu Brain Biosensing

10:35–11:40(65 min)

Overview

Pair up to collect each other's HRV and stress data using Garmin vívoactive 5 and UeduPAD. Analyze multimodal physiological signals in Jupyter Notebooks, including pre-recorded EEG + fNIRS + ECG data from Uedu Brain.

Topics

1
Uedu Fit Hands-on: Paired HRV Collection
Participants pair up—one wears the Garmin watch while the other monitors via UeduPAD, then swap. Analyze each other's stress and heart rate patterns.
2
Uedu Brain Data Lab
Analyze pre-recorded EEG + fNIRS + ECG datasets from Uedu Brain in Jupyter Notebooks, comparing neural signatures across different learning tasks.

Garmin vívoactive 5 & UeduPAD

Participants pair up for this hands-on activity—one person wears the Garmin watch to collect physiological data while the other monitors in real time via UeduPAD iOS App, then they swap roles. Uedu Fit collects the following metrics via the Garmin Connect API:

Heart Rate Variability (HRV) Beat-to-beat interval data reflecting autonomic nervous system activity
Stress Score 0–100 stress index calculated from HRV data
Sleep Stages Four-stage analysis: deep sleep, light sleep, REM, and awake
Body Battery 0–100 energy index combining stress and recovery assessment

HRV Data Collection & Paired Analysis

After collecting paired HRV data, participants analyze each other's stress and heart rate patterns using the Uedu Fit dashboard. Key HRV metrics covered include:

Time-Domain Analysis

SDNN Standard deviation of all NN intervals, reflecting overall HRV
RMSSD Root mean square of successive NN interval differences, reflecting parasympathetic activity

Frequency-Domain Analysis

LF/HF Ratio Sympathetic/parasympathetic balance indicator; ratio increases under stress
Stress Score Garmin's real-time stress score calculated from HRV
Paired Activity: The Complete Collect–Analyze–Interpret Experience

We provide 16 Garmin vívoactive 5 smartwatches. Participants pair up—one wears the watch while the other monitors via UeduPAD, then they swap roles. Those without a watch in a given round use pre-loaded sample datasets.

Uedu Brain Data Lab: Multimodal Biosensing

Uedu Brain is a custom-built wearable multimodal biosensing device that simultaneously measures three physiological signals in a single unit:

EEG (Electroencephalography)

Millisecond-level cortical activity measurement for analyzing attention states, cognitive load, and sleep staging. The team has published a collaborative reasoning framework for edge-deployable EEG sleep staging via local LLM.

fNIRS (Functional Near-Infrared Spectroscopy)

Measures hemodynamic changes in the prefrontal cortex, reflecting brain activation levels during cognitive tasks.

ECG (Electrocardiography)

Research-grade heart rate and HRV measurement with more precise R-R interval data than consumer-grade smartwatches.

In this session, participants use Jupyter Notebooks to analyze pre-recorded multimodal datasets from Uedu Brain, comparing neural signatures across different learning tasks and experiencing EEG sleep staging with the local LLM collaborative reasoning framework.

Related Research

  • Chang, Chen & Juan (2024). Predicting Sports Performance of Elite Female Football Players Through Smart Wearable Measurement Platform. Progress in Brain Research.
  • Cheng, Lin & Chang (2025). Collaborative Reasoning Framework for Edge-Deployable EEG Sleep Staging via Local LLM. IEEE BigDataService.
  • Chu, Garcia & Rani (2023). Research on Wearable Technologies for Learning: A Systematic Review. Frontiers in Education.
  • Sharma & Giannakos (2020). Multimodal Data Capabilities for Learning. BJET.
  • Pijeira-Díaz et al. (2016). Sympathetic Arousal Commonalities and Arousal Contagion During Collaborative Learning. Computers in Human Behavior.

Materials

We will provide 16 Garmin vívoactive 5 smartwatches. Jupyter Notebooks and sample datasets will be prepared.
Tutorial Info
Expected Participants
30
Target Audience
Researchers, educators, and graduate students interested in physiological-aware learning analytics
Prerequisites
Basic understanding of learning analytics concepts
Laptop with modern web browser (macOS or Windows)
No programming experience required (Jupyter notebooks provided)
Invite Us to Your Event

We welcome invitations to bring Uedu workshops to your conference or institution.

Schedule a Meeting