AI for Enhanced Quality of Life
A year-long collaboration led by Dr. Villanueva Rosales with faculty from CS, Civil Engineering, Sociology, Cognitive Science, Geosciences, Education, and Health Sciences is laying the foundation for the Center for AI and Community-Informed Socio-Environmental Research for Healthy Living. The Center aims to build education and research capacity in trustworthy, responsible, and secure human-centered community-driven AI models to address water security and heat risk challenges, focusing on a bi-national, semi-arid region. The subjects of the Center focus specifically on heat risk and amelioration; water security; and AI systems imbued with societal perspectives. The interdisciplinary subprojects will take an integrated approach centered on trustworthy aspects of AI to address economic and environmental challenges that impact healthy living; education and career development that leads to AI-knowledgeable citizens and graduates who have the interdisciplinary, technical, and ethical skills needed for the rapidly changing workforce; and community-informed use cases through participatory processes involving partnerships. The Center’s interdisciplinary faculty team has a record of participatory, user-centered research involving students and the community. The research team has built and continues to build partnerships with community, regional, and national organizations that are essential for engaging the community and achieving the goals of the Center.
Smart, Secure, and Connected Supply-Chain Systems
Led by UTEP with partners from U. Houston, Texas A&M Corpus Christi, UC Santa Cruz, California State U. Dominguez Hills, U. Central Florida, and Kean U., this effort is driven by user needs for AI-enhanced, smart, and secure data systems that consume and generate data and knowledge used for decision-making and will result in resilient, correct, and secure software products and services. In particular, the initiative focuses on the integration of artificial intelligence and cybersecurity. It will involve 20 industry partners in the following areas: Advanced Planning and Predictive Analytics (decision making, demand forecasting, environmental impact monitoring, risk assessment and management); Data and Knowledge Integrity (provenance, integration, and visualization); Data Transactions (blockchain, transaction integrity, safe/secure exchange of information); and Secure Software Products and Services (privacy, security, and efficiency).
Towards Contextualized Data Models on Women’s Health
The robustness and fairness of AI models rely on the quality and diversity of the datasets used for training; however, the demographics, location, or other attributes of the data can impact trust of the results generated by AI models. For example, the NIH All of Us Research Hub, a NAIRR Pilot resource, lacks data from many Healthcare Professional Shortage Areas (HPSA) and from rural communities. Such data sets, when used to address health decisions in rural communities, for example, can lead to conclusions that may be skewed. An emerging initiative, co-led by computer science professor Dr. Natalia Villanueva Rosales and public health science professor Dr. Thenral Mangadu aims to build the human technical infrastructure needed to generate AI-ready datasets representative of underrepresented communities. The project is focused on obtaining community stakeholder input on available and needed regional data specific to women’s health needs; building capacity in communities to adopt data engineering processes (e.g., collection, cleaning, annotation, integration, and publishing) to generate AI-ready data that can lead to trustworthy AI models; and evaluating the use of the data engineering methodologies in the targeted communities to generate AI-ready datasets and leverage NAIRR Pilot resources to advance knowledge in the targeted focus areas. The project plans to extend involvement of health science, social science, public health, computer science researchers, and other stakeholders partnering with communities along the US-Mexico border states, particularly Arizona, New Mexico, and Texas.