From October 11 to 15, 2026, the HCID team led by Professor Ziqian Bai from the School of Automation and Intelligent Manufacturing at Southern University of Science and Technology attended the Ubicomp/ISWC 2026 conference in Shanghai, China. We are pleased to share that our workshop the “Wearable Design in Games and Entertainment” was successfully held on October 11, 2026, as part of UbiComp/ISWC 2026.
Fig1. Group photo of the workshop
A New Design Paradigm for Wearable Game Interaction in the Rehabilitation of People with Limited Hand Mobility
Research Background and Design Framework
Paper “A New Design Paradigm for Wearable Game Interaction in the Rehabilitation of People with Limited Hand Mobility” was accepted by the workshop. This research investigates how wearable sensing technologies and game-based interactions can support rehabilitation for individuals with limited hand mobility. Existing surface electromyography (sEMG)-based rehabilitation systems often prioritize gesture-recognition accuracy while overlooking user needs, wearing comfort, adaptability, and engagement. To address these challenges, we conducted interviews with six individuals with different hand impairments and consulted two rehabilitation therapists to identify key requirements for wearable rehabilitation games. Based on these insights, we proposed a user-centered, four-layer design framework encompassing requirements, technology, conversion, and output, systematically connecting rehabilitation objectives, wearable sensing capabilities, gesture recognition, and game interaction design.
Fig2. Paper presentation and report
Fig3. The design process of the fishing-themed rehabilitation game
System implementation, experimental results and future work
To demonstrate the framework's feasibility, we developed a fishing-themed rehabilitation game incorporating a comfortable, reusable textile-electrode armband for collecting forearm muscle signals. The system translates three fundamental hand gestures—fist clenching, pinching, and finger extension—into interactive fishing actions, enabling users to control a virtual fishing rod and catch fish through muscle activation. By integrating real-time biofeedback, adaptive challenges, and engaging game mechanics, the prototype explores how rehabilitation exercises can be transformed into more motivating interactive experiences. Preliminary experiments with healthy participants achieved approximately 98% in-game gesture-recognition accuracy under controlled conditions, demonstrating the technical feasibility of the proposed approach.
Future work will focus on clinical validation with individuals with limited hand mobility and long-term evaluation in real rehabilitation settings. This study highlights the potential of combining wearable computing, human-centered design, and gamification to create more accessible, adaptive, and engaging rehabilitation experiences.
Fig4. Participants' sharing and game experience zone
Authors
Ziyan Xia¹, Zhexi Wang¹, Zeyu Yang², Yucong Cai², Yuxuan Zhong², Tengjia Zuo², Christina Zong-Hao Ma³, Xiaoqian Deng⁴, Kaiyi Qiu⁴, Ziqian Bai¹
¹ Southern University of Science and Technology
² The Hong Kong University of Science and Technology (Guangzhou)
³ The Hong Kong Polytechnic University
⁴ Guangdong Work Injury Rehabilitation Hospital
NapScarf Presented at UbiComp/ISWC 2026
On the afternoon of October 12, 2026, student from HCID presented NapScarf, an intelligent neck pillow for commute naps, at the Intelligent Soft Wearables workshop at UbiComp/ISWC 2026, held at the Shanghai International Convention Center. Through a presentation and prototype demonstration, she introduced the project, designed and developed under the supervision of Professor Bai Ziqian. NapScarf combines soft sensing, adaptive pneumatic support, and on-body vibration reminders to address unstable head and neck support and concern about missing a stop when napping on public transport.
Fig5. Inside structure and interactive mode
Research Background
Napping on public transport offers a chance to rest, but vehicle motion and concern about missing a stop can disrupt the experience. Conventional neck pillows provide limited adjustment as head posture changes, while phone vibrations or station announcements may not reliably reach a sleeping passenger. NapScarf addresses these challenges by combining posture sensing, adaptive pneumatic support, and private vibration reminders in a soft neck pillow, exploring a wearable approach to supporting rest during everyday commutes.
Interaction Prototype Design
NapScarf integrates three pressure sensors and independently controlled airbags around the left, right, and rear of the neck. When sustained pressure changes indicate a head tilt, the corresponding airbag inflates to support the head and neck and guide the head toward a neutral position. The system also provides an on-body vibration reminder, reducing reliance on public announcements or continuous physical contact with a phone. In the current prototype, this reminder is triggered manually.
An initial user study suggested that participants valued the prototype’s head and neck support and the privacy of its reminders. The study also identified opportunities to improve wearing comfort, reduce pump noise, and further integrate the hardware.
Presentation and Hands-on Demonstration
During the presentation, student introduced the commute nap scenario, NapScarf’s layered sensing and airbag structure, and the interaction between head posture and pneumatic support. In the demonstration that followed, attendees gathered around the display to examine the prototype, learn how it worked, and try on the pillow to explore its fit and support.
The on-site exchanges extended beyond technical implementation to consider the experience of using NapScarf in everyday life. They brought attention to design questions such as fit across different users, comfort during prolonged wear, and how vibration reminders might be perceived and accepted on public transport. These exchanges also offered directions for further development, including improving fit and adjustability, reducing bulk and noise, and making support and reminders a more natural part of the resting experience.
Fig6. Student introduces the NapScarf concept and system design during the workshop presentation
Fig7. Student is introducing the prototype of NapScarf
Fig8. An attendee is trying on NapScarf
Authors
Peicen Zhao¹, Ziqian Bai¹
¹ Southern University of Science and Technology
From Stigma to Support: Designing a Low-stigma Smart Knee Pad for Community-based KOA Exercise Management
At UbiComp/ISWC 2026 in Shanghai, HCID team presented From Stigma to Support: Designing a Low-stigma Smart Knee Pad for Community-based KOA Exercise Management in the Design Exhibition. The paper looks at older adults who manage knee osteoarthritis through community exercise. Moderate activity helps the knee, yet many seniors misjudge the load of a single session. A hospital look keeps clinical monitors off the body in daily life because of stigma, so physicians find it difficult to obtain accurate information about the daily knee joint movements of elderly individuals.
Prototype Design and Workflow
The prototype is a low-stigma knee pad, co-designed with senior exercisers and community physicians for community exercise with knee osteoarthritis. It combines a silicone patella cushion, a neoprene body, and adjustable Velcro straps. A Seeed XIAO nRF52840 Sense sits under the cushion, and its LSM6DS3 IMU streams six-axis motion at 50 Hz over Bluetooth. The overall interaction workflow is as follows:
1.Before exercise: The first session asks the wearer to stop when mild knee discomfort is first felt. The personal limit is set just below the Knee Motion Index (KMI) reached at that point. KMI accumulates flexion energy and dynamic acceleration, and each later session is read against this personal limit.
2.During exercise: The phone classifies the movement and, during walking or running, flags an out-toe gait. A progress bar, with voice and text, warns at 80% and 100% of the limit.
3.After exercise: The wearer rates fatigue and knee status. These ratings, read with the KMI trace, inform the next limit. A physician view sums the week, marks days past the limit, and shares activity and gait notes for community exercise advice.
Throughout the process, protection, sensing, and feedback stay on one sports-form object, so daily exercise keeps a clear personal boundary.
The team also designed a companion app for this loop. Users complete calibration, follow the in-session cues, and review a short report after exercise.
Fig9. The Prototype in Design Exhibition
Discussion and Learning
Beside the demo, we discussed the work with scholars from the Hong Kong University of Science and Technology (Guangzhou), Harbin Engineering University, and other universities. They affirmed the co-design process and the age-friendly concept of destigmatization. They also suggested a next sensing step—adding further sensors, including electromyography—so muscle effort can sit alongside the motion-based KMI. We brought that direction back as a concrete path for the next prototype.
Fig10. On-site conversation around the leg-worn exhibition.
Authors
Chengwei Wu¹, Linyue Sun¹, Qi Fang¹, Ziqian Bai¹
¹ Southern University of Science and Technology
Somni: An AI Companion System for Intervening in Bedtime Procrastination among Older Adults
In October 2026, the HCID team from Southern University of Science and Technology presented Somni: An AI Companion System for Intervening in Bedtime Procrastination among Older Adults at the AAA Workshop of UbiComp/ISWC 2026. The project explores how natural, noncoercive interactions can help older adults reduce excessive smartphone use before bedtime and develop healthier routines.
Prototype Design and Workflow
Somni combines a physical companion robot with a mobile application. After users upload a personal photo, a multimodal large language model and AudioLDM 2 work together to generate a semantically matched sleep-aid soundscape. Before bedtime, users simply place their phones on the robot’s sensing area to trigger voice feedback and audio playback. The app also supports sleep logging and a lightweight star-reward mechanism.
Fig11. The interaction workflow and Somni prototype.
Fig12. Some team members pose with the Somni poster at UbiComp/ISWC 2026.
Discussion and Learning
In an exploratory usability study with seven participants, feedback highlighted the clear interaction flow and greater acceptability of phone placement compared with forced screen locking. The photo-based soundscape feature was also well received, while screen readability and audio richness were identified as areas for improvement.
During discussions at the workshop, we received a suggestion to explore millimeter-wave radar for simple, contactless detection of whether a user has fallen asleep. This could potentially reduce reliance on wearable devices such as smartwatches and offers a new direction for extending Somni’s sleep-sensing and automated companionship capabilities.
Authors
Yajing Li¹, Diandian Sun¹, Ziqian Bai¹
¹ Southern University of Science and Technology
