Shu Kit Eric Tam – Assistant Professor of Neuroscience (Research Field: Animal behavior and neuroscience)

shukiteric.tam@dukekunshan.edu.cn

Brief description of the project: LumiHabit is a simple “animal-centric” interactive system that will allow small pets (e.g., hamsters, mice, and gerbils) to self-administer light and darkness in their enclosure based on their innate preference. This system will use Skinnerian conditioning principles (Tam S.K.E. et al., 2025, Journal of Behavioral Addictions)-either via lever pressing, noses poking, or wheel running-to let the animal control ambient light and darkness. This will allow an individual animal to select specific wavelength within the visible light spectrum, brightness, and duration in its home cage. The objective is to let your pet (rather than the pet owner) to be in control of its own lighting ennvironment, so that it can be matched closely with its sleep/wake cycle and behavioral states across the 24-h day. This has the potential to promote the pet’s wellbeing as well as prolonging its lifespan.

Expected outcome(s): Students will actualize the concept of LumniHabit-developing a prototype of the device (with a data acquisition component) that can be validated and optimized in small pets.

Claudia Nisa – Assistant Professor of Behavioral Science (Research Field: behavioral science; health promotion)

claudia.nisa@dukekunshan.edu.cn

Brief description of the project: This project aims to produce and disseminate scientifically-driven audiovisual materials (WeChat Mini Program) for tobacco control in Jiangsu Province, in collaboration with Kunshan and Jiangsu Health Commissions. The main objective is to target (1) pregnant women, their partners and their parents during gestation via the 413 Maternity and Child Hospitals in Jiangsu Province and (2) parents and grandparents of infants up to 3 years old in Pediatrics services in the 18 Community Health Centers in Kunshan. Videos and messages to be disseminated in a WeChat Mini Program will be theoretically grounded in Behavioral Science, culturally adapted to China norms and habits and allow data collection from the population (e.g., survey; data usage).

Expected outcome(s):
Report for the Kunshan and Jiangsu Health Commissions
Two peer-reviewed papers
Reduce smoking rates during pregnancy by 20%
Reduce smoking rates during infancy by 10%

Claudia Nisa – Assistant Professor of Behavioral Science (Research Field: Behavioral Science)

claudia.nisa@dukekunshan.edu.cn

Brief description of the project:
The PAUSE and SEAT initiative is a forward-thinking project designed to encourage a healthier and more mindful commuting experience by promoting smartphone-free seating. By restricting the use of smartphones in designated seating areas within the Shanghai subway system, we aim to foster detachment from constant digital engagement while encouraging passengers to rest, relax, and recharge during their commute.
Modern urban living has led to excessive dependence on smartphones, often resulting in fatigue, stress, and a lack of connection with the immediate environment. The PAUSE and SEAT program offers commuters a compensatory benefit: those who choose to abide by this guideline will have access to seats, with the added value of participating in a collective effort to reduce tech-induced strain. Those wishing to continue using their smartphones will have to stand up.
Implementation involves 1) designating seating areas as smartphone-free zones through clear signage, 2) public announcements, and visual cues in the seating area. Subway staff or AI-powered monitors can gently enforce compliance with non-intrusive reminders. The project is paired with a public awareness campaign highlighting the importance of digital breaks for mental health.
By providing this unique seating option, the Shanghai subway system can position itself as a global pioneer in promoting healthier and more mindful urban commuting habits.

Expected outcome(s):
1.Research Report: A comprehensive analysis documenting the development process, user insights, data analysis, findings, and recommendations for future improvements.
2.Published Research Paper: Submission and acceptance of peer-reviewed research papers in relevant journals or conferences, disseminating the findings to the academic and professional community.
3.Proof-of-Concept (PoC) or Prototype: A functional prototype or working demonstration clearly illustrating the core idea and its feasibility, ready for further iteration or scaling.

Bing Luo – Assistant Professor of Data and Computational Science (Research Field: LLM based AI Systems)

bing.luo@dukekunshan.edu.cn

Brief description of the project:
ChatDKU (https://chatdku.dukekunshan.edu.cn/) is an AI project at Duke Kunshan University, aiming to build a RAG-Agent chatbot system that provides smarter academic and campus services.
ChatDKU leverages a retrieval-augmented generation (RAG) pipeline and a memory-enhanced cyclic agent architecture to provide contextually relevant and precise information. Unlike traditional AI systems that rely on less secure cloud-based or API-based solutions, ChatDKU features that all data and AI models are hosted securely on DKU’s local servers, ensuring robust data security and privacy.
ChatDKU has successfully launched V1.0 in collaboration with Advising and Signature Work Offices, we are looking for self-motivated and skilled students to exploit and add new features and functionalities to improve user experience in ChatDKU.

Expected outcome(s):
Students participating in the ChatDKU project will have the opportunity to engage in a multifaceted learning experience that bridges cutting-edge research with practical engineering. Responsibilities include enhancing the agentic RAG architecture, fine-tuning and quantizing language models, improving backend and frontend systems, and implementing robust DevOps workflows. Students will also analyze user feedback and usage data, develop tools for telemetry analysis, and evaluate the quality of ChatDKU’s responses using state-of-the-art frameworks.
This project offers significant educational benefits, equipping students with expertise in AI development, natural language processing, and system design. By tackling challenges such as prompt engineering, structured output generation, and user feedback analysis, students will refine their problem-solving and critical thinking skills. The work provides a holistic view of deploying AI solutions in real-world scenarios, including considerations for scalability, maintainability, and user experience.
Through hands-on contributions, students will enhance their programming and analytical capabilities while gaining exposure to advanced AI techniques like Retrieval-Augmented Generation, LLM fine-tuning, and automated evaluation frameworks. The anticipated outcomes-system improvements, live demonstrations, and potential publications or patents-will significantly bolster students’ academic profiles and prepare them for professional opportunities in artificial intelligence and software engineering.

Yiu wing Kam – Assistant Professor of Global Health (Research Field: Global Health)

yiuwing.kam@dukekunshan.edu.cn

Brief description of the project:
The primary objective of this project is to develop a novel HPV detection and diagnosis platform to accelerate cervical cancer elimination efforts in alignment with the WHO 90-70-90 targets. By increasing access to high-performance screening, enabling early and accurate detection of high-risk HPV types, and ensuring timely treatment, the project aims to reduce cervical cancer incidence and mortality. Specifically, it focuses on improving screening coverage to at least 70% of women by the ages of 35 and 45, while creating scalable and sustainable solutions for underserved and resource-limited settings.
Theoretically, this project offers insights into the application of advanced diagnostic technologies in global health, particularly in areas where infrastructure and resources are limited. It also contributes to understanding the spatial epidemiology of HPV prevalence and the socio-demographic determinants of cervical cancer in low- and middle-income countries.
Practically, the project demonstrates how cutting-edge technologies such as point-of-care testing and cloud-based data systems can be integrated into existing healthcare frameworks to achieve large-scale health targets. The lessons learned from this project could inform future implementations of advanced diagnostics for other diseases, from a logistical, financial, and clinical perspective.
The project exhibits innovation by utilizing a next-generation HPV detection platform that combines portability, affordability, and advanced AI diagnostic capabilities. Unlike traditional lab-based HPV testing, the proposed platform supports same-day diagnosis and triage, reducing delays in screening and follow-up care. A key innovation lies in its real-time data connectivity, which allows centralized monitoring of screening programs, identification of missed populations, and better targeting of vaccination campaigns. Moreover, its integration with mobile clinics and community health workers ensures equitable access to marginalized populations, addressing systemic gaps in cervical cancer prevention and treatment.
Key collaborations between the public sector (DKU) and the private sector (a precision diagnostic company) will play a crucial role in ensuring the project’s feasibility.

Expected outcome(s):
The expected outcomes could include the following, depending on the project’s scope and objectives:
•A detailed research report summarizing findings and insights
•At least one accepted or published research paper in a peer-reviewed journal or conference
•Development of intellectual property (e.g., patents or trademarks) for innovative solutions or technologies
•Enhanced collaborations and networks for future research and development

Lijun Liu (Senior Coordinator for DKU Innovation Incubator) – co-mentor with Dr. Liqi Ren, (Senior Lecturer of Innovation and Design, Associate Director for Innovation and Entrepreneurship） (Research Field: sociology; social inclusion; sustainability; inclusive employment; public policy) lijun.liu@dukekunshan.edu.cn

Brief description of the project:
This project is a social innovation lab designed to move beyond academic observation and into community-based action. We aim to understand and support families of young adults with intellectual disabilities during their crucial transition to employment.
1.Academic Value & Practical Significance:
Academic Value: This project addresses a significant research gap by shifting the focus from resource-scarce regions to the unique challenges within a prosperous area like Jiangsu. It offers a valuable comparative perspective and will generate novel, localized data on family resilience.
Practical Significance: The findings will provide direct, evidence-based insights for NGOs, social service organizations, and local policymakers. The primary output will be a tangible, co-designed workshop toolkit that can be immediately implemented to improve family support systems, contributing directly to the regional goals of inclusive employment and common prosperity.
2.Innovation:
The core innovation is methodological. We are not just studying a community; we are building with them. By employing an Action Research framework and facilitating a series of social innovation workshops, we merge rigorous data collection with hands-on, collaborative problem-solving.
This interdisciplinary approach combines sociological inquiry with human-centered design principles, focusing on co-designing and prototyping solutions with families, rather than for them.
Well-Defined Research Focus:
3.The project is tightly anchored by four clear, sequential research questions that form the backbone of our inquiry:
RQ1: What are the primary challenges and stressors for these families?
RQ2: How do their family resilience processes (belief systems, organization, communication) function in this context?
RQ3: What specific coping strategies do these families actively employ?
RQ4: Based on this data, can we identify distinct, localized models of family adaptation?
4.Feasible Research Plan:
The project is structured into a realistic, three-phase plan designed for an undergraduate team to complete within one academic year:
Phase 1 – Foundation & Outreach (Months 1-3): Systematically review existing literature and build trusted relationships with local partner NGOs and initial family participants in the Kunshan/Suzhou area.
Phase 2 – Data Collection & Co-Design Workshops (Months 4-8): Conduct semi-structured interviews and surveys, while simultaneously running a series of interactive social innovation workshops to gather qualitative data and collaboratively prototype support strategies.
Phase 3 – Analysis & Toolkit Creation (Months 9-12): Analyze all collected data to answer the core research questions and synthesize the findings and workshop outcomes into the final “Family Empowerment Workshop Toolkit.”

Expected outcome(s):
This project is designed to generate both rigorous academic insights and tangible, community-focused outputs:
1.Primary Outcome: A Comprehensive Research Report: A detailed report that answers our four core research questions, providing an evidence-based analysis of family resilience in the Jiangsu region.
2.Academic Paper: A manuscript prepared for submission to a relevant undergraduate research conference, detailing our unique mixed-methods approach and findings.
3.Prototype of a Family Empowerment Workshop Toolkit: This is the core social innovation output. It will be a well-documented, replicable toolkit-including facilitation guides, activity templates, and case studies-based on the solutions co-designed and validated during our workshops.
4.A short documentary or multimedia case study: A compelling visual story documenting the workshop process and the experiences of the participating families, perfect for sharing the project’s impact with a wider audience.

Yucheng Jin – Assistant Professor of Computer Science (Research Field: Human-centered computing for well-being and creativity)

yucheng.jin@dukekunshan.edu.cn

Brief description of the project:
Music is a powerful medium for emotional expression and social connection, yet Deaf and Hard-of-Hearing (DHH) individuals often face barriers to experiencing it due to its reliance on auditory perception. Recent AI-driven systems have attempted to bridge this gap by combining music information retrieval (MIR) with generative visual synthesis for audio-reactive displays. While these approaches provide visual channels for experiencing music, they still struggle to convey the expressive and dynamic qualities of music in a more embodied and socially engaging way.
We propose a low-latency robotic-arm-driven music visualization system that transforms audio features into UV-light trajectories projected onto a phosphorescent surface. Unlike conventional virtual visualizations, the robotic arm introduces a tangible, embodied dimension to music by transforming auditory signals into kinetic motion patterns that can be visually perceived and physically felt. At the same time, the integration of light visualization adds an additional layer of multimodal feedback, mapping musical features such as rhythm, loudness, and timbre into dynamic visual cues.
This approach builds on the idea that music is not only heard, but also seen and felt. By extending musical interaction through robotic motion and light embodiment, our system aims to create more inclusive and immersive experiences that enable DHH individuals not only to personally connect with music, but also to share in social practices such as dancing with others-whether in pairs, at concerts, in parks, or even across generations. In doing so, the system fosters a stronger sense of belonging, enabling DHH individuals to engage in musical and social worlds from which they have historically been excluded.

Expected outcome(s):
A demo, an art exhibition, and a research paper

Luyao Zhang – Assistant Professor of Economics (Research Field: Interdisciplinary study of computational and economics science on groundbreaking technologies including big data, blockchain, generative ai, geospatial science, and innovative computing)

luyao.zhang@dukekunshan.edu.cn

Brief description of the project:
In 1994, Harsanyi, Nash, and Selten received the Nobel Prize in Economics for pioneering game theory, transforming our understanding of strategy. However, its abstract mathematical demands limit accessibility, and its potential for digital transformation-where platforms, auctions, and data-driven competition dominate-remains underexplored. “Play to Smart” aims to develop a gamified educational app that immerses students in high-stakes economic and business decision-making scenarios-such as mergers and acquisitions, auctions, competitive bidding in big-tech platforms, and strategic voting within corporate governance. Through rich, AI-adaptive game mechanics grounded in game theory and behavioral economics, students will research, simulate, and innovate while exploring real-world digital-economy dilemmas, ensuring that each playthrough challenges learners to think critically and creatively about strategic decision-making.
From a computational and design standpoint, the project will involve:
• Developing AI-driven adaptive scenarios that modulate complexity and feedback based on player strategies.
• Crafting an engaging UI/UX front-end using principles from media, art, and game aesthetics-storyboarding, character design, data visualization, and interactive narrative.
• Leveraging generative AI tools for content variation, scenario generation, and adaptive storytelling.
In doing so, “Play to Smart” aligns with SDGs:
• Quality Education (SDG 4)-by offering immersive, decision-centric learning.
• Industry, Innovation, and Infrastructure (SDG 9)-by integrating AI and digital tools.
• Peace, Justice, and Strong Institutions (SDG 16)-by exploring governance and collective strategic choices.

Expected outcome(s):
• Research deliverables: Detailed research report aiming at publications at international conferences.
• Proof-of-concept: Fully playable app showcasing at least one high-stakes scenario.
• Applied outreach: Live demo at industry events, participation in innovation hackathons/contests, path for incubation or entrepreneurial exploration.

Ziv Ze’ev Cohen – Assistant Professor of Digital Media (Research Field: Human-Machine interface)

zivzeev.cohen@dukekunshan.edu.cn

Brief description of the project:
Tactile feeling is conventionally defined as a sensory experience produced by direct physical contact, involving the activation of mechanoreceptors in the skin. Touch has been central to how humans interact with their environment, with its role extending to our engagement with tools, instruments, and machines.
In contemporary contexts, however, we are increasingly interacting with multiple machines simultaneously. The field of Human-Computer Interaction (HCI) has therefore become a critical site for development, experimentation, and research. As computational devices and interactive systems become more mundane and ubiquitous, they no longer remain external tools but instead shape our behaviors, perceptions, and social relations in profound ways.
Traditionally, interaction has relied heavily on tactile input and reciprocal feedback (e.g., pressing a button, swiping a screen). Such exchanges often form two-way loops between human action and machine response, practices in which users grow increasingly proficient over time.
This project, however, seeks to extend this inquiry by focusing on Spatial Interaction and Contactless HCI-forms of interaction that do not require physical touch. Instead, they rely on gesture recognition, body movement, or spatial presence in 3D environments. The aim of this project is to investigate the effect generates in the absence of tactile input, and how such interactions are cognitively and socially processed.
Although tactile feeling is generally linked to physical contact, there are many well-documented phenomena where people experience touch-like sensations without direct touch. Examples include phantom sensations, such as those associated with phantom limb experiences; haptic illusions produced by carefully designed visual or auditory cues; synesthesia, where one sensory modality evokes the perception of another; and cross-modal perception, in which the integration of senses produces tactile impressions in the absence of physical contact. These cases suggest that tactile feeling is rooted neurologically and perceptually.
This project will produce a creative experience supported by an academic text where these aspects can be examined first-hand, allowing viewers/users to perceive and reflect on the complexities of contactless tactility. By situating audiences within an environment that deliberately blurs the boundary between physical touch and its perceptual substitutes, the project is calling for an experience that provokes a deeper discussion about the nature of sensation, embodiment, and the shifting conditions of Human-Computer Interaction.

Expected outcome(s):
The project will take the form of a creative artwork that integrates both physical and digital components. While the primary focus is on the artwork itself, it may also be supported by a research paper, though this will remain a secondary priority.
The aim is to develop an artwork that can be submitted to a range of international events, bridging both the fields of Art & Technology and Human-Computer Interaction (HCI). By positioning the work within these overlapping domains, the project seeks to contribute both as an artistic expression and as a design-driven exploration of interaction and experience.

Zhenyu Yang – Primary: Visiting Assistant Professor of Physics and Computer Science; Secondary: Research Assistant Professor under the Medical Physics program (Research Field: Medical Physics, Radiation Therapy, Cancer Treatment)

zhenyu.yang893@dukekunshan.edu.cn

Brief description of the project:
This project investigates the environmental impact of radiotherapy departments across different healthcare systems, focusing on Duke University Medical Center (USA) and Kunshan First People’s Hospital (China). It introduces a pioneering, data-driven framework to benchmark energy use, clinical waste, and institutional sustainability practices in radiation oncology. Utilizing clinical audits, carbon footprint modeling, and structured governance analysis, this comparative study aims to generate actionable insights for building greener cancer treatment systems.
In addition to its academic value, the project emphasizes innovation and entrepreneurship by proposing the development of a student-led consultancy prototype, which will translate our research findings into practical sustainability solutions. This “Sustainable Oncology Lab” consulting unit will offer hospitals and radiotherapy centers customized audits, carbon modeling services, and workflow recommendations. The consulting model will serve as a foundational step toward entrepreneurship incubation, giving students experiential learning in sustainability assessment, stakeholder engagement, and healthtech service design.
The project is moderately challenging, with clearly defined goals in data collection, environmental modeling, and policy translation. It will employ student researchers in equipment audits, interview transcription, and comparative analysis, fostering transdisciplinary problem-solving and entrepreneurial thinking.
Phase 1: Baseline Audit (Month 1-3)
• Conduct field audits at Kunshan First People’s Hospital to collect data on equipment usage (LINACs, CT simulators), energy consumption patterns, and waste streams.
• Develop standardized data collection templates for student assistants, including energy logging sheets and consumables inventory forms.
• Interview clinical staff and administrators to understand local sustainability practices and bottlenecks.
Phase 2: Comparative Modeling (Month 4-6)
• Build carbon footprint models for different treatment pathways (e.g., hypofractionated vs. conventional radiotherapy) using patient flow, equipment runtime, and transportation data.
• Benchmark these findings against open-access datasets or published metrics from Duke University Medical Center.
• Train students in basic life-cycle assessment (LCA) tools and Excel-based emissions calculators.
Phase 3: Consultancy Prototype Development (Month 7-9)
• Translate findings into a structured consultancy toolkit, including: A clinical sustainability scoring rubric; PowerPoint-based audit report templates; Custom recommendations for energy saving and waste reduction
• Pilot this consultancy model through a mock audit session or student-led presentation to hospital stakeholders.
Phase 4: Dissemination and Iteration (Month 10-12)
• Prepare student posters, conference abstracts, and/or research summary brochures.
• Solicit feedback from clinicians and DKU faculty to refine service offerings.
• Explore the feasibility of formalizing the “Sustainable Oncology Lab” as a student-run service unit within DKU Innovation Incubator.

Expected outcome(s):
• A cross-national dataset on sustainability metrics in radiation oncology
• Research reports and peer-reviewed manuscripts (e.g., Advances in Radiation Oncology, Medical Physics)
• Development of a carbon benchmarking tool and consultancy prototype (PoC)
• Prototype of a consultancy service model for environmental audits in radiotherapy
• Undergraduate student capstone presentations and poster dissemination
• Policy recommendations for clinical administrators and sustainability officers

Fujie Xu – Full Professor of Global Health; Co-Director of Global Health (Research Field: Global Health / Mental Health)

fujie.xu@dukekunshan.edu.cn

Brief description of the project:
(1) Executive Summary
Undergraduate students frequently encounter chronic stress, fatigue, and performance pressure during their studies-factors that compromise both mental health and academic success. A recent report from Harvard’s Graduate School of Education highlights that nearly 27% of students already experience burnout, while 52% face multiple sources of stress, including academic expectations, social norms, and future planning pressures. Yet, universities still lack effective tools to capture and monitor the overall mental well-being of their student body in a systematic way. PROMIS is intended as a voluntary, dual-purpose platform that empowers students with personalized self-insight-tracking well-being indicators like sleep, workload, and stress-while giving the university anonymized, aggregate data to guide targeted support.
(2) Problem Statement
Undergraduate students frequently experience chronic stress caused by high academic workload and insufficient sleep, both of which are strongly linked to reduced concentration, impaired memory, lower academic performance, and increased risk of burnout and mental health issues:
• Sleep deprivation impairs concentration and memory, reducing learning outcomes (Hershner & Chervin, 2014).
• Chronic stress increases the risk of burnout, depression, and attrition (Dyrbye et al., 2005).
• Universities lack systematic, real-time insights into the well-being of their students, relying instead on sporadic surveys or self-reporting in crisis situations. For institutions valuing innovation, global competitiveness, and student well-being, there is a clear chance for a scalable, research-driven tool to monitor and support resilience across the student body.
(3) Solution: Platform for Resilience, Optimization, and Mental-health Insights for Students (PROMIS)
The PROMIS is supposed to be a dual-purpose digital platform designed for undergraduates and their institution:
• For students:
◦ Voluntary weekly check-ins with person-ralted data-input on workload, sleep quality, perceived stress, and recovery habits.
◦ Personal dashboard that visualizes individual trends and provides actionable recommendations (e.g., schedule adjustments, recovery activities, mental health resources).
• For the university (aggregate level only):
◦ Institutional view (aggregate level, with consent): By default, the institution receives only anonymized, aggregated data to monitor student well-being trends. If students opt in, they can share specific personal data with university medical or counseling services, enabling timely, personalized support and intervention.
◦ Data-driven recommendations for targeted interventions (e.g., outreach to specific cohorts, resilience workshops, UG policy adjustments).
Key innovation: Unlike standard wellness apps, the PROMIS links individual resilience building with institutional monitoring, ensuring both students and their institution benefit.
(4) Research Study Plan & Focus
To ensure the PROMIS reflects the needs of DKU students, we will conduct a baseline study in semester one (Session 2, Fall 2025):
• Participants: 200 DKU undergraduate students balanced regarding their demographics (voluntary, anonymous).
• Methods:
◦ Online survey (expectations, willingness to use dashboard, perceived stress factors).
◦ Focus groups for qualitative insights into workload, sleep, and recovery habits.
• Research focus:
The research will identify measurable indicators of student well-being (e.g., sleep, stress, workload, recovery habits) as well as culturally relevant parameters, assess students’ willingness to share personal data under voluntary and anonymous conditions, and develop a culturally contextualized algorithm to generate actionable, personalized recommendations. The findings will directly inform the design of the PROMIS dashboard and guide university-level strategies for supporting student mental health.
• Research goals:
◦ Identify the key metrics to track mental well-being.
◦ Understand students’ willingness to share data for health-related purposes.
◦ Develop an algorithm that evaluates the collected data to provide individualized recommendations for students (culturally and contextually relevant).
(5) One-Year Timeline
Phase 1 (Months 1-2): Research & Design
• Conduct baseline survey and focus groups at DKU.
• Finalize metrics and algorithm design.
• UX mock-ups of the dashboard.
Phase 2 (Months 3-6): Prototype Development
• Build MVP (web-based dashboard).
• Create student and institutional interfaces.
• Internal testing with a small pilot group (n=30).
Phase 3 (Months 7-9): Pilot Rollout
• Launch voluntary participation for 150-200 students.
• Collect data, monitor dashboard use.
• Iterate design based on user feedback.
Phase 4 (Months 10-12): Market Readiness Evaluation
• Analyze collected data (impact on awareness, willingness, trends).
• Refine prototype and algorithm.
• Deliver final report.

Expected outcome(s):
• Research: Comprehensive dataset on DKU undergraduate resilience and well-being.
• Prototype: Working dashboard MVP (minimum viable product) with student and institutional views.
• Entrepreneurship: Pitch deck and business model (SaaS) for international scaling.
• Impact: Tangible contribution in DKU’s capacity to support student well-being while strengthening global reputation as a pioneer in mental health on campus.
In sum: research report, prototype, intellectual property, business model