2026 ASEE Annual Conference & Exposition

MACHINE LEARNING APPROACHES TO FORECAST HOUSEHOLD TELEHEALTH ADOPTION: A DATA GROUNDED FRAMEWORK

Presented at DSAI-Session 12: Data Science Projects, Datasets, and Real-World Applications

Household telehealth adoption remains uneven, particularly in communities facing limited broadband access and low levels of digital readiness. Traditional analytic approaches often describe adoption patterns retrospectively and frequently overlook the structural conditions that shape who is able to engage with telehealth services. In the absence of theory-driven, data-grounded frameworks, planners, organizations, and public agencies lack effective tools to support outreach strategies, infrastructure investment, and equitable resource allocation. As telehealth becomes an increasingly essential component of healthcare access, understanding the conditions that shape household-level adoption is critical for ensuring equitable service delivery. This study adopts a data-grounded, theory-guided design to examine household telehealth adoption, with an emphasis on feasibility assessment and framework development rather than fitted model performance. Guided ex ante by the Unified Theory of Acceptance and Use of Technology (UTAUT), the study specifies how facilitating conditions, effort expectancy, social influence, and performance expectancy can be operationalized using observable indicators related to broadband access, digital readiness, demographic context, and health needs. The analysis draws on empirically observed distributions from community-based survey data and supporting public-sector information, supplemented by qualitative insights that contextualize barriers to telehealth use. This paper contributes a theory-constrained, data-grounded framework that integrates technology adoption theory with exploratory analytic techniques to structure empirical signals related to access, digital readiness, and perceived usefulness at the household level. Rather than training or validating supervised learning models, the study evaluates whether the available data exhibit sufficient structure to support the future application of supervised machine learning within a theory-constrained analytic space. Observed patterns are interpreted at the level of UTAUT constructs to inform scenario-based reasoning about adoption readiness, rather than to generate population-level predictions. The resulting framework provides a transparent and policy-relevant foundation for targeted planning, guiding future data collection efforts and supporting equitable telehealth strategies in underserved communities. In addition, the framework contributes to engineering education by providing a transparent, theory-driven example of how data, machine learning concepts, and equity considerations can be integrated and taught in applied, data-constrained contexts.

Authors

Samuel Fakolade Morgan State University [biography]

Samuel Fakolade is a PhD researcher at Morgan State University, Maryland, specializing in civil engineering, infrastructure systems, and transportation engineering, with interdisciplinary research interests in digital equity and public service systems. His work focuses on digital health equity, telehealth systems, broadband accessibility, and technology adoption in underserved communities. His research examines household telehealth adoption, broadband-enabled healthcare access, and digital laptop adoption disparities in HBCU anchor communities. He uses data-driven approaches and systematic review methods to study how infrastructure and connectivity influence access to healthcare and education. With a professional background in structural engineering, infrastructure inspection, and construction project coordination, he combines practical engineering experience with policy-focused research. His broader academic interests include healthcare accessibility, digital inclusion, and equitable infrastructure development.
Ms. Obiageli Gertrude Adaku Nwachukwu Morgan State University [biography]

Obiageli Nwachukwu is a Graduate Research Assistant at the Centre for Equitable Artificial Intelligence and Machine Learning Systems, Morgan State University, where she specializes in data‑driven modeling, machine learning, and applied bioinformatics. Her research spans predictive analytics, digital health adoption, and the development of equitable AI frameworks for underserved communities. She has contributed to NTIA‑funded projects focused on broadband access, telehealth readiness, and community‑centered digital equity initiatives. With a background in pharmacy and advanced training in computer science (bioinformatics), she integrates domain expertise in health systems with computational methods to build predictive models that support informed decision‑making in public health and technology adoption. Her broader interests include clinical trial outcome prediction, translational data science, and AI‑enabled solutions for improving health access and outcomes.
Julius Ogaga Etuke http://orcid.org/0000-0002-2710-2803 Morgan State University [biography]

Julius Ogaga Etuke is a seasoned civil engineer and Ph.D. Candidate in Civil Engineering at Morgan State University, where his research focuses on sustainable and resilient infrastructure engineering. With over 15 years of consulting and project leadership experience, primarily in Nigeria's infrastructure landscape, he brings a robust, practical perspective to academic inquiry. His doctoral research is at the cutting edge of transportation resilience, integrating virtual and augmented reality (VR/AR), climate adaptation modeling, and data-driven solutions to improve disaster preparedness and risk mitigation in complex transportation systems. As a COREN-registered engineer and member of the Nigerian Society of Engineers, his work emphasizes impactful research and consultancy that bridges structural design principles with urgent sustainability and resilience challenges.
Dr. Oludare Adegbola Owolabi P.E. Morgan State University [biography]

Dr. Oludare Owolabi, a professional engineer in Maryland, joined the Morgan State University faculty in 2010. He is the director of the Sustainable Infrastructure Development, Smart Innovation and Resilient Engineering Research Lab, Department of Civil Engineering at Morgan State University
Dr. Uttam Gaulee http://orcid.org/0000-0002-3010-2812
Dr. Olushola V. Emiola-Owolabi http://orcid.org/https://0000-0002-8521-5769 Morgan State University [biography]

Olushola Emiola-Owolabi is an engineering pedagogy specialist, researching active learning in engineering classrooms and specializing in qualitative methods research on teaching and learning – particularly in remote synchronous learning environments.
Olorunfunmi Samuel Shobowale Morgan State University [biography]

Olorunfunmi Shobowale is a graduate research assistant at Morgan State University, Maryland, pursuing a Masters degree in Advanced Computing. He has over nine years of experience in IT, edtech and STEM education. His research focuses on equitable engagement strategies and the integration of AI in learning environments.
Russel Patrick Chanza Morgan State University [biography]

Russel Chanza is a Graduate Research Assistant at Morgan State University, where he is pursuing a Master of Engineering in Civil Engineering. He is a member of the Sustainable Infrastructure Development, Smart Innovation, and Resilient Engineering (SID-SIRE) Research Laboratory, where he engages in systematic data organization, analysis, and interpretation, contributing to the development of rigorous, policy-ready research deliverables. His research examines broadband accessibility, digital equity, and equitable infrastructure systems, with a particular focus on how connectivity gaps and socioeconomic determinants influence technology adoption among underserved populations. His academic interests are centered on applying machine learning, artificial intelligence, and predictive modeling to translate empirical findings into actionable, equity-informed policy guidance for digital public services.

Note

The full paper will be available to logged in and registered conference attendees once the conference starts on June 21, 2026, and to all visitors after the conference ends on June 24, 2026

« View session