Texas A&M University ยท Industrial & Systems Engineering
PhD student in Industrial and Systems Engineering, serving as a graduate research and teaching assistant at Texas A&M University.
This website is where I share my research, publications, and academic work.
Overview
I study advanced statistical learning methods for quality, reliability, and prognostics. My interests include Bayesian modeling, Gaussian processes, survival analysis, physics-informed neural networks, and Bayesian last-layer neural networks.
I care about models that capture the behavior and constraints of real engineering systems, communicate uncertainty clearly, and remain useful to decision-makers in noisy, real-world environments. That balance between rigor and usefulness defines the kind of research I want to build.
Focus Areas
Failure prediction and survival analysis for engineering systems.
Bayesian hierarchical modeling, scalable Bayesian neural networks, and uncertainty quantification for prediction and decision support.
Combining physical laws with machine learning under noisy or limited data.
Featured work
This paper was recently published in Technometrics (2026) and was a finalist for the QCRE Best Student Paper Competition at the 2025 IISE Annual Conference & Expo.
Abstract
Modern industrial systems are often subject to multiple failure modes, and their conditions are monitored by multiple sensors, generating multiple time-series signals. Additionally, time-to-failure data are commonly available. Accurately predicting a system's remaining useful life (RUL) requires effectively leveraging multi-sensor time-series data alongside multi-mode failure event data. In most existing models, failure modes and RUL prediction are performed independently, ignoring the inherent relationship between these two tasks. Some models integrate multiple failure modes and event prediction using black-box machine learning approaches, which lack statistical rigor and cannot characterize the inherent uncertainty in the model and data. This paper introduces a unified approach to jointly model the multi-sensor time-series data and failure time concerning multiple failure modes. This proposed model integrates a Cox proportional hazards model, a Convolved Multi-output Gaussian Process, and multinomial failure mode distributions in a hierarchical Bayesian framework with corresponding priors, enabling accurate prediction with robust uncertainty quantification. Posterior distributions are effectively obtained by Variational Bayes, and prediction is performed with Monte Carlo sampling. The advantages of the proposed model are validated through extensive numerical and case studies with a jet-engine dataset.
View publicationQuick access
Dive into themes, methods, and selected work.
Browse papers, work in review, and manuscripts in preparation.
See a structured overview of education, experience, and publications.
Reach out for academic, research, or industry conversations.
Browse event photos, conference highlights, and presentation snapshots.