Initiatives Cohort

Lead PI: Soon Lee

Team: Soon Lee, Jiho Noh, Dylan Goldblatt

Project Summary: Our research develops an AI-powered tool that helps science teachers improve how they lead classroom discussions, a teaching practice proven to deepen student understanding and close achievement gaps. Unlike commercial AI systems that fail to recognize when students are genuinely reasoning through scientific problems versus simply recalling facts, our Automated Discourse Analysis System (ADAS) has achieved 68-79% accuracy in identifying these critical learning moments. This seed funding has enabled us to build a working prototype, validate our approach with expert educators, and assemble the foundation for a full NSF proposal that aims to put research-grade professional development tools in the hands of every science teacher.

Lead PI: Razvan Voicu

Team: Razvan Voicu, Muhammad Hassan Tanveer, Yanique Tello, Ayse Tekes, Dylan Goldblatt

Project Summary:

Lead PI: Zongxing Xie

Team: Zongxing Xie, Chen Zhao, Xinyue Zhang, Mark Geil

Project Summary: Our research team is developing a system that can monitor how people walk (their “gait”) in their everyday environments, without needing them to wear any devices. This type of continuous, contactless monitoring can help detect early signs of health issues such as fall risks, mobility decline, or the onset of neurodegenerative conditions like dementia, as the individuals walk about their own life. By using advanced radar sensors and AI technologies, our project aims to make in-home health monitoring more accessible, affordable, and reliable. This research brings us closer to creating smarter homes that can help families and healthcare providers intervene earlier and more effectively, supporting aging individuals and improving overall quality of life. With support from the seed grant, we successfully developed a working prototype for real-world gait data collection and analysis. We also completed a scoping review paper -- accepted to HumanSys 2025, The Third International Workshop on Human-Centered Sensing, Modeling, and Intelligent Systems Co-located with Cyber-Physical Systems and Internet-of-Things Week 2025 -- that surveys the current landscape of wireless gait sensing research and highlights key opportunities and challenges in the field. In addition, we engaged with technical and health care experts through a survey study to better understand end-users’ perspectives on adopting smart sensing technologies in caregiving practices for individuals with Alzheimer’s disease and related dementias (ADRD). Through this project, we engaged and trained three undergraduate students as research assistants, providing them with hands-on experience in sensing technologies, data analysis, and human-centered research. This opportunity helped prepare them for future careers or graduate studies in computing, healthcare technology, and interdisciplinary research.

Lead PI: Awatef Ergai

Team: Awatef Ergai, Sylvia Bhattacharya
 
 Project Summary: To enhance the quality of hospital care and improve healthcare delivery, researchers at Kennesaw State University are designing a personalized, AI-powered task assistance system. The first phase of this project employed a user-centered design approach, engaging directly with patients and healthcare providers to better understand their needs, preferences, and concerns related to developing generative AI tools for healthcare settings. Participants expressed openness to the use of digital technologies in hospital settings, particularly when tools are intuitive, informative, and supported by clear communication. Across perspectives, there was strong emphasis on emotional support, personalized care, and assistance with tasks such as understanding health information, managing medications, streamlining documentation, supporting clinical decision-making, safeguarding privacy, and preserving the human connection in care. The seed grant enabled this foundational work, ensuring the future AI-powered system is designed with users, not just for them. By aligning technology with real-world needs, this research aims to enhance patient outcomes, reduce provider burnout, and foster more compassionate, efficient, and responsive hospital care.
 

Lead PI: Alice Gooding

Team: Alice Gooding, Da Hu
 
Project Summary: At Kennesaw State University, we are using advanced simulation techniques to improve how ground-penetrating radar (GPR) detects buried human remains. By simulating realistic burial scenarios and analyzing GPR signal responses, our research helps forensic teams better interpret subsurface data and identify potential grave sites with greater accuracy. Supported by the Interdisciplinary Seed Grant, this project has produced valuable insights that enhance training, guide future field applications, and ultimately support faster, more effective investigations. This work demonstrates KSU’s role in advancing technologies that directly benefit public safety and justice.

Lead PI: Geza Kogler

Team: Geza Kogler, Ayse Tekes, Coskun Tekes

Project Summary: Bioinspired, variable-stiffness joints offer a promising pathway to improve gait performance and user satisfaction by better matching the mechanical characteristics of the orthosis to individual joint behavior. Such designs have the potential to reduce fatigue, restore more natural gait dynamics, and ultimately support greater independence and quality of life for individuals with foot drop. Many of today’s orthotic and pediatric prosthetic devices are built using rigid materials and standard designs that don’t adapt to the individual. Therefore, they often feel uncomfortable, don’t match how joints naturally move, and can lead to poor walking patterns over time. Even though we have advanced tools in robotics and prosthetics, similar innovation hasn’t reached orthotics, especially not in a way that’s affordable, lightweight, or easy to customize. Adults with foot drop often face increased risk of falling, reduced walking efficiency, and decreased physical activity due to the mechanical limitations of standard orthotic devices.

The project focused on two aims:

1. Design and development of a compliant and variable stiffness joint to replicate passive ankle stiffness observed during walking, aiming to support more efficient and natural gait in individuals with foot drop.

2. Development of an ankle stiffness measurement test device.

Passive Ankle Foot Orthosis Joints (pFOA). We designed several bio-inspired and compliant passive joints that replicate the nonlinear stiffness behavior of the human ankle, as shown in Figure 2.

Preliminary Compliant AFO Design. Our initial design, as illustrated in Fig. 3, consists of a fully foot plate, a compliant joint, and the shank section that extended along the posterior calf.

Ankle joint stiffness measurement apparatus. We designed an ankle joint stiffness measurement apparatus (as shown in Fig. 4) to conduct “normative” baseline measurements of ankle joint stiffness on healthy subjects with different foot lengths using a custom-built ankle joint stiffness measurement system.The ankle stiffness measurement system uses a computer-controlled custom motorized system that dorsiflexes the foot ankle quantifying ankle joint angle. It also simultaneously measures and records the force (torque) resistance that results during dorsiflexion range. We also developed a custom graphical user-interface based Matlab application which can control the hardware and perform recording of the experimental data as shown in Fig. 5.

Lead PI: Bo Li

Team: Bo Li, Ashish Aphale, Beibei Jiang, Lei Shi

Project Summary: The goal is to leverage directional supramolecular assembly of oligopeptides to build 3D architecture with interconnected conductive network, tunable mechanical properties and hierarchical structures over multiple length scales, for flexible electrode materials towards applications in solar cells and batteries. Active electrode materials with 3D architecture for flexible solar cells and batteries share a common requirement: sufficient pathways for ion/molecule transport as well as conductive network for electrons to exit into external circuits. However, it has proven to be extremely difficult to build such 3D architecture with molecular resolution using conventional 2D lithographic fabrication methods (i.e. “top-down” strategy). To this end, crafting 3D architecture via self-assembly of functional nanomaterials using a “bottom-up” paradigm provides a promising solution to these technological challenges. Among various nano-building blocks, biomolecules such as DNA and peptide exhibit promising capability as building blocks for constructing 3D architecture with nanoscale resolution, due to their programmable intermolecular bonding. The proposed work will provide a feasible strategy for fabrication of 3D architecture by controlling supramolecular assembly of oligopeptides at multi-length scale. The proposed work will use a combination of experimental and analytical methods to understand the 3D architecture formation process, and, more importantly, how such process determines its electrical and mechanical properties of the formed structures.

Lead PI: Melissa Osborne

Team: Melissa Osborne, Lei Zhang, Allison Garefino

Project Summary: Injury is the leading cause of death among children in the U.S., and the home environment is a common location where these injuries occur. While educational programming and information campaigns exist to help parents better understand how to keep their children safe at home, people learn best when they can get hands-on practice. Simply reading information may help, but it doesn’t always result in behavior change. Immersive virtual reality (IVR) allows individuals to practice skills in a risk-free environment. With support from the KSU Office of Research Interdisciplinary Seed Grant, our team developed an IVR home safety program called ParentSHIELD-VR. In ParentSHIELD-VR, parents enter a virtual home environment, identify child safety hazards and address them, and manage child supervision in the midst of simulated distractions. We heard from real parents to find out what they wanted to learn about home safety and their thoughts about interacting with IVR. We used their input to develop ParentSHIELD-VR and then tested it with another group of parents. From these test sessions, we learned more about what parents think of the program, how it is helping them learn home safety skills, and what we can improve. We are using this data to take our project to the next level for further development and testing. There are few things parents want more than to keep their children safe and secure. Through ParentSHIELD-VR, we are helping to equip parents with not only the knowledge, but the practice and confidence they need to create safe spaces for their children.

Lead PI: Herman Ray

Team: Herman Ray, Volker Franke

Project Summary: This pilot project uses a new type of artificial intelligence (AI) to look at how fairly resources and opportunities are shared among high schools in Cobb County, Georgia. The goal is to understand whether all students, no matter their background, are being treated equitably.

To do this, we use a tool called a "knowledge graph" and include human oversight to guide the process. The AI helps us organize and analyze information from documents like student handbooks, school improvement plans, staff assignments, and available student services. These documents are often written in different formats, so we use AI to find patterns that might otherwise be hard to spot.

By comparing information across different schools in the same district, we look for differences in how equity policies are written.

Lead PI: Razvan Voicu

Team: Razvan Voicu, Muhammad Hassan Tanveer, Yannique Tello

Project Summary: The WESHARE (Wheelchair Exoskeleton Synergy for Holistic Adaptive Rehabilitation Evolution) project at Kennesaw State University is pioneering a new standard for how mobility-impaired individuals receive daily rehabilitation. Millions of people around the world rely on wheelchairs, yet traditional rehabilitation methods often require them to travel to clinics and use complex equipment that may not be accessible, affordable, or tailored to their needs. WESHARE aims to change that by transforming the wheelchair into an intelligent, multifunctional rehabilitation platform.

This project combines a lightweight, modular lower-limb exoskeleton with a high-functionality wheelchair frame to support seated users in performing essential physical therapy exercises at home or in care facilities. Whether users are recovering from injury, stroke, or managing a lifelong condition, the system offers continuous therapy through passive, assistive, and resistive training modes. It specifically targets critical muscle groups in the legs and feet to improve blood circulation, prevent muscle atrophy, and maintain joint flexibility—all while users remain safely seated. This daily movement is especially important for older adults and rural populations who may face barriers to consistent in-person therapy.

What makes WESHARE unique is its combination of advanced robotics and human-centered design. The system will integrate real-time sensors and intelligent control algorithms to ensure movements are safe, adaptive, and aligned with the user’s unique biomechanics for human trials. A planned modular mounting system allows the exoskeleton components to attach and detach easily to most wheelchair models, increasing accessibility. Future versions will include integrated health monitoring tools such as heart rate, oxygen level, and blood pressure sensors, along with telemedicine features so healthcare providers can assess patients remotely.

The WESHARE team developed a fully operational prototype, conducted early testing, and presented findings at multiple academic conferences. These efforts have laid the groundwork for expanded research, commercialization, and improved access to rehabilitative care. This work represents KSU’s commitment to research that matters, enabling greater independence, better health outcomes, and higher quality of life for people with mobility challenges.

Lead PI: Lei Zhang

Team: Lei Zhang, Anisah Bagasra

Project Summary: Bereaved parents who have lost a child to cancer often report a feeling of isolation and a loss of social connectedness after leaving the treatment hospitals. Many of them also struggle to make sense of and find meaning in their traumatic experience of loss due to the untimely death of a young child. Challenges with finding meaning have been associated with prolonged and complicated grief symptoms that can adversely affect grieving parents' quality of life. Although existing videoconferencing and social media tools can be used to provide bereavement support remotely, their user experiences are generally not engaging because of the fourth wall effect (an imaginary barrier that separates an audience from the fictional/virtual world), and they lack custom features to support meaning-centered grief therapies. Leveraging the unique affordances of multi-user immersive virtual reality (IVR) system and text-to-image generative artificial intelligence (AI), we hypothesize that an immersive and individualized therapeutic virtual experience, with support of meaning-centered grief techniques will have great potential fostering social connectedness in bereaved parents in shared virtual worlds and impact their healing journey positively. This seed grant project serves as the preliminary work to the proposed main project with focuses on the following objectives: 1. To assess the user acceptance of the technology proposed by the main project with bereaved parents and grief counsellors. 2. To develop a risk management protocol with grief counsellors and evaluate it. 3. To develop a proof-of-concept multi-user IVR prototype. Outcomes from this seed grant project helped the PI and his team successfully achieved two goals: 1. A conference paper published to Mixed/Augmented Reality for Mental Health (MARMH) at 2024 International Symposium on Mixed and Augmented Reality (ISMAR). 2. Secured an NSF CRII grant award (https://www.nsf.gov/awardsearch/showAward?AWD_ID=2451461&HistoricalAwards=false) for the main project.