Upcoming Event: Oden Institute Seminar

PIER (Physics-Informed, Energy-efficient, Risk-aware routing): Toward Practical, Risk-Aware Voyage Planning

Aniruddha Bora, Texas State University

Abstract

Ships often encounter short, steep waves that make voyages uncomfortable, slow, and fuel hungry. Many routing tools still optimize mainly for arrival time and only sketch the sea state. This talk presents PIER, a practical routing framework that blends physics insight with machine learning to plan routes that are fast, energy-efficient, and risk-aware. Our approach integrates a fast risk indicator, a compact performance surrogate, and an offline reinforcement-learning policy inside a rule-aware routing engine to balance comfort, energy, and time while preserving safety constraints. Using simulated and public-data scenarios, I will discuss representative case studies and preliminary comparisons, with ablation studies to identify the most influential signals. The work culminates in a portable sandbox suitable for independent evaluation and deployment.

Biography

Aniruddha Bora is an Assistant Professor in the Department of Computer Science at Texas State University. He earned his Ph.D. in Computational Analysis and Modeling from Louisiana Tech University under the supervision of Dr. Weizhong Dai. Before joining Texas State University, he was a Postdoctoral Research Associate in the Division of Applied Mathematics at Brown University with Dr. George Karniadakis. His research focuses on Scientific Machine Learning (Physics-Informed Neural Networks and Neural Operators), and Hybrid Methodologies (Numerical Solver + Machine Learning) for complex multiscale physical systems. In particular, he develops novel neural-operator frameworks, hybrid numerical–machine learning solvers, and multi-fidelity operator approaches, with applications in turbulence, extreme events, nanoscale heat transfer, and metamaterials. He is also actively working in interpretable machine learning, aiming to build models that not only achieve high predictive accuracy but also provide insights into the underlying physical and statistical mechanisms. Dr. Bora’s contributions have appeared in prestigious venues such as Advanced Materials; Proceedings of the Royal Society A; International Journal of Heat and Mass Transfer; Applied Mathematics and Computation; Communications in Computational Physics; Neural Networks; and AAAI. His recent co-authored work includes an ICLR 2025 CCAI Workshop paper and an AI4X conference paper on explainable-AI frameworks for extreme weather. He also serves the scientific community as an external reviewer for leading journals and conferences in machine learning, scientific computing, and applied mathematics, and he is an ATPESC alumnus.