Exploring Health Through Data, Science and Innovation
Bio360AI is a comprehensive Health Digital Twin architecture that unifies longitudinal multi-modal datasets with predictive computational engines. Bio360AI began with a question about recognizing changes in health earlier. The work progressed from a multimodal Health Digital Twin framework to two IEEE publications and a mobile application designed to make longitudinal health insights more accessible.
Dr.Bing Ma
Assited in comparing microbial changes associated with five immunosuppressants across three time points.
Presented a comparative analysis evaluating whether SparCC networks and heatmap-based methods identified consistent microbial patterns across immunosuppressants and time points.
Sun2Water
Solar-Powered Water Purification
Developing a scalable, off-grid system for producing clean water using solar energy, multistage filtration, and UV treatment.
Mentor: Dr. Srijib Mukherjee | 2025–2026
KEY PROJECT CONTRIBUTIONS
Developed a solar-powered water purification system combining multistage filtration, UV treatment, and digital performance monitoring.
Designed the system to produce approximately 300–800 liters of purified water per day.
Built and tested a working prototype to evaluate filtration performance, energy requirements, and potential scalability.
Examined how an off-grid purification system could improve access to clean water in communities with limited infrastructure.
RECOGNITIONS
Top 50 among more than 1,200 national submissions in the Mott Million Dollar Challenge.
First Place in the One Montgomery Green Sustainability Challenge.
Second Place at the Marriott Sustainability Summit.
Recognized as a Top 50 project in the Green Inventors Challenge.
Received more than $2,000 in project funding and support.
10+ Participants | Exercise + EEG | Individualized Reports
LLM Evaluation | NSF-Supported Research | Software Security
BRAIN BOOST
EXERCISE AND COGNITIVE HEALTH
UMBC RISE LAB
Mentor: Tarannum Shaila Zaman, Ph.D
Compared physical activity with cognitive and neurological indicators.
Identified participant-level patterns across exercise and EEG data.
Developed individualized reports summarizing each participant’s results.
Compared model performance across detection, explanation, and repair tasks.
Evaluated AI-generated fixes for accuracy and reliability.
Contributed to the research paper and collaborated with a multidisciplinary team.
