Wearable AI & Sensing

#smartwatches #PPG #IMU #LightSensing #SleepSensing #PhysiologicalSensing #EdgeSensing #SensorFusion #FoundationModels #labStreamingLayer #LSL

Core Question: How can consumer wearable/IoT devices capture meaningful representations of human physiology and behavior in everyday life?

🏆Funded by JSPS KAKENHI

📈 Highly cited publications (Top 10%)

📈 Publications in JMIR mHealth uHealth, JHIR

🏅 Honorable Mention for Best Presentation Award @ JSSR 2021, 2022

Wearable Sleep Monitoring and and Sleep Architecture Analysis

Advances in consumer wearable devices have created new opportunities for continuous sleep monitoring outside clinical settings. This research investigates the validity of wearable technologies, including smartwatch-based activity trackers and wearable EEG devices, for measuring sleep duration, sleep architecture, and sleep stage transitions in real-world environments. We also develop machine learning approaches to improve sleep stage prediction using wearable sensor data, enabling more accurate and accessible sleep assessment without traditional polysomnography. By evaluating both the capabilities and limitations of consumer sleep technologies, this work contributes to the development of reliable, scalable, and user-friendly solutions for long-term sleep monitoring and digital sleep health.
💻 GitHub Repo

Selected Publications

[Q1] Liang Z, Chapa-Martell MA. (2021) A multi-level classification approach for sleep stage prediction with processed data derived from consumer wearable activity trackers. Frontiers in Digital Health (Section: Health Informatics) 3:665946. Doi: 10.3389/fdgth.2021.665946. [PubMed /Scopus] (Highly cited top 14%)
[Q1, top 10%] Liang Z, Chapa-Martell MA. (2019) Accuracy of Fitbit wristbands in measuring sleep stage transitions and the effect of user-specific factors. JMIR mHealth and uHealth 7(6):e13384, DOI:10.2196/13384. (Highly cited top 6%)
[Q1] Liang Z, Chapa-Martell MA. (2018) Validity of consumer activity wristbands and wearable EEG for measuring overall sleep parameters and sleep structure in free-living conditions. Journal of Healthcare Informatics Research 2 (1-2): 152-178. (Highly cited top 7%)
Sirithummarak P, Liang Z. (2021) Investigating the effect of feature distribution shift on the performance of sleep stage classification with consumer sleep trackers. In Proceedings of the IEEE 10th Global Conference on Consumer Electronics (GCCE 2021), Kyoto, Japan. 【First author is master student】[Scopus ]

🏆Funded by JSPS KAKENHI

🤝Collaborations with Tokushima University (徳島大学)

🏅 Best Presentation Award @ WSSE 2024

Realtime Multimodal Data Streaming from Smartwatch

Consumer smartwatches have evolved into powerful sensing platforms capable of continuously capturing physiological data in everyday environments. This research develops software platforms, middleware, and machine learning techniques that enable reliable, high-resolution acquisition, streaming, and analysis of wearable sensor data in real time. Our work spans wearable-based sleep stage prediction, middleware for low-latency physiological data streaming, and tools that integrate smartwatch data with interactive applications such as game engines and digital health systems. By bridging wearable hardware, software engineering, and artificial intelligence, this research provides the technological foundation for next-generation ubiquitous computing, digital phenotyping, and personalized healthcare applications.
💻 GitHub Repo

Selected Publications

Liang Z. WearStreamer: Enabling High-Resolution Health Data Acquisition from Consumer Smartwatches. In Proceedings of the 7th World Symposium on Software Engineering (WSSE 2025), Okayama, Japan. [Scopus]
Liang Z. (2024) Middleware Solution for Real-Time Physiological Data Streaming from Consumer Smartwatches. In Proceedings of the 6th World Symposium on Software Engineering (WSSE 2024), Kyoto, Japan. (reviewed)【Best Presentation Award】
Liang Z. (2024) Harnessing Consumer Smartwatches for Heart Rate Streaming in Unity Game Engine. In Proceedings of 2024 IEEE Consumer Life Tech Conference (ICLT 2024), Sydney, Australia. [Scopus]
Sirithummarak P, Liang Z. (2023) Developing a Cross-Platform Application for Integrating Real-time Time-series Data from Multiple Wearable Sensors. In Proceedings of the 10th International Electronic Conference on Sensors and Applications (ECSA2023), Cyberspace.【First author is master student】[Scopus]

🏆 Supported by KUAS Cornerstone Project

🏆 Supported by iCAN

🏆 Supported by MEMS Park Consortium

🏅 2nd Place @ iCAN’26 Contest

🏅Finalist in upcoming Ubicomp 2026 Student Challenge

🏅Finalist in upcoming Student Competition in Microsystems Applications

ThermalGuardian: An Assistive Wearable System for Home-Based Monitoring and Prevention of Bathing Related Health Risks

Bathing is an essential daily activity but can pose significant health risks, particularly for older adults and individuals with chronic medical conditions. This project develops ThermalGuardian, a wearable monitoring system that continuously tracks physiological responses during bathing to detect early signs of heat-related stress and other potential health risks. By integrating wearable sensors with intelligent risk assessment algorithms, the system provides real-time monitoring and timely alerts to promote safer bathing practices in home environments. The project aims to support independent living, reduce bathing-related accidents, and demonstrate how wearable technologies can enable proactive, personalized healthcare in everyday life.
🌐Official Website 🐼GitHub Repo 🕹️Contest 📰News 📣Instagram

Selected Publications

Koo J, Toendan KA, Shimada ET, Akaiduzzaman AKM, Liang Z. Thermal Guardian: An Assistive Wearable System for Home-Based Monitoring and Prevention of Bathing Related Health Risks in Older Adults. In Proceedings of the 19th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2026), Marbella, Spain.【First author is undergrad student】 (reviewed)

🏆 Supported by KUAS Cornerstone Project

🏆 Supported by iCAN

🏅 Special Award for Technology @ iCAN’26 Contest

KAKI: A Modular Smart Insole Platform for Health, Mobility, and Human Performance

The feet provide rich physiological and biomechanical information that can reveal how people walk, balance, and interact with their environment. This research develops next-generation smart insole technologies for continuous lower-body sensing in everyday life. Our work focuses on designing modular, reusable, and customizable wearable platforms that integrate plantar pressure sensors, inertial measurement units (IMUs), and environmental sensors into a flexible sensing architecture. By combining novel hardware design, additive manufacturing, embedded systems, and intelligent signal processing, we enable accurate monitoring of gait, foot pressure distribution, posture, and movement patterns while reducing cost and electronic waste through modular electronics. Beyond hardware development, we investigate how multimodal insole data can support applications such as gait analysis, rehabilitation, fall-risk assessment, indoor pedestrian navigation, and personalized health monitoring. Our long-term vision is to establish smart insoles as an unobtrusive, sustainable, and scalable wearable platform for continuous lower-body sensing, advancing consumer digital health, human movement analysis, and ubiquitous computing.

Selected Publications

Coming soon...

🏆 Supported by KUAS Cornerstone Project

🏅 Winning Team @ TALE 2020 Student Research Competition (Undergraduate Category)

Akikomi: Smart Occupancy Sensing with Grid-Eye Ultrasound Sensor Arrays

Understanding how shared spaces are used is essential for improving comfort, energy efficiency, and facility management in smart buildings. This project developed AkiKomi, a mobile sensing system that estimates room occupancy in real time using data collected from smartphones and other readily available devices. By providing accurate, privacy-conscious occupancy information without requiring specialized infrastructure, the system enables more efficient utilization of classrooms, offices, and public spaces. The project demonstrates how ubiquitous computing and mobile sensing technologies can support intelligent building management while enhancing the everyday experience of occupants.

Selected Publications

Tanaka D, Tanaka J, Ikuru K, Liang Z, Piumarta I. (2022) AkiKomi: Design and implementation of a mobile app system for real-time room occupancy estimation. In Proceedings of the 2022 IEEE 11th Global Conference on Consumer Electronics (GCCE 2022), Osaka, Japan. [Scopus]【First authors are undergraduate students】

🏆 Supported by KUAS Cornerstone Project

🏅 Finalist Team @ iCAN’25 Contest

From Skin to Screen: Smart Bandage for Real-Time Wound Monitoring

Effective wound care requires regular assessment, yet frequent clinical visits can be inconvenient and visual inspection may not detect complications at an early stage. This project develops a smart bandage system that integrates wearable sensors with a smartphone application for continuous, real-time wound monitoring. The system captures indicators related to the wound environment and presents the data through an accessible mobile interface, enabling users and caregivers to track healing progress and identify potential risks. By combining flexible sensing, wireless communication, and mobile health technologies, the project aims to support proactive wound management, timely intervention, and more accessible home-based healthcare.

Selected Publications

Hossain FMM, Tan DJG, Mohamed IA, Ayun QQ, Matsunaga S, Junaideen AS, Liang Z. (2025) From Skin to Screen: Real-Time Wound Monitoring with a Smart Bandage and Smartphone. In the Proceedings of the 2025 IEEE 14th Global Conference on Consumer Electronics (GCCE 2025), Osaka. Japan.【First author is undergraduate student】[Scopus]

🏆 Supported by KUAS Cornerstone Project

🤝 International Collaborations with OSU (USA)

Inclusive PPG: Adaptive Heart-Rate Monitoring Across Diverse Skin Tones

Photoplethysmography (PPG) is widely used in smartwatches and wearable health devices, yet its accuracy can vary across users because skin tone, sensor placement, movement, and environmental conditions affect signal quality. This project develops an adaptive PPG sensing system designed to provide more reliable heart-rate monitoring across diverse populations. By combining multiple optical wavelengths, intelligent signal-quality assessment, and personalized machine learning, the system dynamically selects or integrates the most informative sensor signals for each user. The project aims to reduce measurement bias in wearable technologies and contribute to more inclusive, equitable, and clinically dependable digital health monitoring.

Selected Publications

Coming soon...

Page updated

Google Sites

Report abuse