2026 ASEE Annual Conference & Exposition

ReefCast for Coral Resilience: A Research-based Interdisciplinary Project for Developing a Machine Learning Model for Predicting Coral Bleaching

Presented at Ocean and Marine Division (OMED) Technical Session 1

Coral reefs, which support nearly 25% of all marine species, are facing a global crisis due to climate change-induced mass bleaching events. These vital ecosystems also protect coastlines from storms and erosion and support economic activity through fisheries and tourism, making their preservation a global imperative. Across the world, traditional forecasting models rely largely on Sea Surface Temperature (SST) and Degree Heating Weeks (DHW), however, these statistical methods struggle to capture the complex, non-linear interplay of multiple environmental stressors such as ocean acidification, turbidity, and salinity changes that trigger these devastating ecological events.

Current approaches in engineering education are increasingly tackling such multifaceted problems through interdisciplinary frameworks. Many programs utilise project-based learning (PBL) and senior capstone projects that require students to build solutions by integrating principles from engineering, data science, and environmental biology, while also emphasising research-based methodologies that promote scientific inquiry and data-driven problem solving. However, even with these pedagogical advancements, environmental challenges like coral bleaching are often overlooked in undergraduate research.

This paper presents ReefCast, a research-driven interdisciplinary project focused on building a solution to this global problem. The project was born out of an integration of two courses: Integrated Communication for Engineering [ICE for Engineering [ICE for Research]Engineering (ICE) and Machine Learning and Pattern Recognition [MLPR].).]. This research-guided approach to machine learning proved effective for developing an AI/ML model that could predict coral bleaching. Our approach began with intensive research into the problem of coral bleaching, which consisted of exploring traditional solutions, studying the evolution of machine learning models used within the domain, and developing our solution. The research principles taught in the ICE course helped us find often-neglected oceanic and coral features and incorporate them as features of our model while exploring novel approaches for its development.

This initial phase served as the project's proof of concept, requiring the intensive integration of a 25-year, multi-source dataset and the implementation of methods to handle the severe class imbalance typical of rare event data. This provided a rich, end-to-end engineering challenge, pushing us to move beyond simple accuracy, which is often misleading in imbalanced datasets, evaluating performance using a comprehensive suite of metrics where recall was prioritized, as the ecological cost of a false negative (a missed bleaching event) is far more damaging than a false positive. This rigorous evaluation demonstrates that our machine learning models can significantly outperform traditional approaches, as our models were engineered to process the complex, non-linear interactions between multiple stressors, allowing them to identify bleaching events that simpler methods might miss.

Beyond technical innovation, this project serves as a case study in engineering education for marine and coastal ecosystem health, demonstrating how research-based and project-oriented pedagogy can bridge theory and practice. Students learned to balance accuracy with ecological responsibility, selecting evaluation strategies that prioritise the prevention of missed bleaching events. ReefCast thus represents both a scientific contribution to coral reef conservation and is evidence of the importance of using research-based interdisciplinary projects for training future engineers to design solutions to global environmental challenges.

Authors

Dr. Brainerd Prince Plaksha University [biography]

Brainerd Prince is the Associate Professor and the Director of the Center for Thinking, Language and Communication at Plaksha University. He teaches courses such as Reimagining Technology and Society, Ethics of Technological Innovation, and Art of Thinking for undergraduate engineering students and Research Design for PhD scholars. He completed his PhD on Sri Aurobindo’s Integral Philosophy from OCMS, Oxford – Middlesex University, London. He was formerly a Research Tutor at OCMS, Oxford, and formerly a Research Fellow at the Oxford Centre for Hindu Studies, a Recognized Independent Centre of Oxford University. He is also the Founding Director of Samvada International Research Institute which offers consultancy services to institutions of research and higher education around the world on designing research tracks, research teaching and research projects. His first book The Integral Philosophy of Aurobindo: Hermeneutics and the Study of Religion was published by Routledge, Oxon in 2017. For more information, please visit: https://plaksha.edu.in/faculty-details/dr-brainerd-prince
Dr. SIDDHARTH SIDDHARTH Plaksha University [biography]

Dr. Siddharth leads the Human-Technology Interaction Lab at Plaksha University. He recently came to Plaksha from the Human Factors Center of Excellence at Microsoft USA where he worked as Senior Human Factors Researcher. He is passionate about working on research themes with practical applications in the technology industry.

At Plaksha, Dr. Siddharth is investigating and modeling human physiological responses to improve cognitive well-being, and enhance human safety, productivity, and creativity. By integrating domain-specific knowledge from the fields of signal processing, computational neuroscience, computer vision, and machine learning, his research group is utilizing human physiological responses to track various cognitive states such as emotions, stress, awareness, mental load, and more. His research and recommendations have practical applications in the technology industry for the development of safer and more user-centered designs that promote cognitive well-being.

In the past, Dr. Siddharth led the Microsoft Human Factors Center of Excellence’s efforts to design personalized user experiences through neurophysiological insights into users' preferences enabling them to be more productive and establish meaningful connections with each other. He has also worked at Facebook Reality Labs, Samsung Research America, French National Center for Scientific Research (CNRS), and National University of Singapore. His research has been awarded fellowships and research grants from various institutions including the National Science Foundation, Army Research Lab, and the Kavli Institute of Brain and Mind. It has also been published in many international conferences and reputed journals including IEEE Transactions and Nature Scientific Reports and has spawned multiple patents.

Dr. Siddharth received his Ph.D. from the University of California San Diego where he also served as the Graduate Fellow Editor of the Prospect Journal of International Affairs and Graduate Student Member of the South Asia Initiative. His first book, Founding Generations: Democracy's Origins and Parallels in America and India, was released recently. For more up-to-date information about him, please visit https://ssiddharth.in/about-me
Rhythm Mehra Plaksha University [biography]

Rhythm Mehra is an undergraduate student in Biological Systems Engineering at Plaksha University, focused on building data-driven solutions for health and environmental sustainability. Her work sits at the intersection of biology, machine learning and systems engineering, with an emphasis on predictive modeling and real-world impact. She combines computational thinking with hands-on lab experience, having trained in microbiology and fermentation at Orbit Biotech, enabling her to bridge experimental biology with scalable, data-centric solutions.
Rishit Anand Plaksha University [biography]

Rishit Anand is an undergraduate student of Computer Science and Artificial Intelligence at Plaksha University, specializing in machine learning, reinforcement learning, and data engineering. His current research interests span applied machine learning, human-technological interaction, and the alignment of human values in Large Language Models.
Shrijak Kumar Plaksha University [biography]

Shrijak Kumar is an undergraduate student in Computer Science and Artificial Intelligence at Plaksha University, focused on building AI infrastructure, systems, and software for real-world impact. His work spans end-to-end machine learning pipelines and data-intensive systems. He has industry experience as a Machine Learning and Software Engineering Intern at Vibracoustic and as a Research Associate at the University of Missouri. He also served as CTO at Personhood, where he built scalable AI systems for autonomous agents and real-time applications.

Download paper (469 KB)

Are you a researcher? Would you like to cite this paper? Visit the ASEE document repository at peer.asee.org for more tools and easy citations.

» Download paper

« View session