About

Devin Harris leads the Infrastructure Simulation, Sensing and Evaluation Lab (I-S²EE Lab) at the University of Virginia. The lab focuses on developing advanced techniques to characterize and better evaluate the condition of the built environment. Their research is interdisciplinary in nature, maintaining a core emphasis on assessing the performance and condition state of infrastructure using a variety of methods. While much of the research concentrates on transportation infrastructure, the lab also explores related domains such as smart and connected communities, cyber-physical systems, crowd-sourcing, and smart/high-performance materials. The physical laboratory is housed within the Department of Civil and Environmental Engineering and is equipped with testing equipment for rapid load testing, monitoring, and non-destructive evaluation of civil infrastructure, supporting both laboratory and field evaluations.

Research topics

• Computer Science
• Machine Learning
• Artificial Intelligence
• Data Mining
• Algorithm
• Data science
• Database

Selected publications

• What Lies within: Utilizing Graph Neural Networks for Subsurface Detection in Finite Element Simulations

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access
  PublisherDOI

• What lies within: Utilizing graph neural networks for subsurface detection in finite element simulations

  Engineering Structures · 2025-06-23 · 4 citations

  article
  PublisherDOI

• Iterative Updating of Digital Twins Using Convolutional Neural Networks: A Framework for Robust Structural Behavior Prediction

  2025-06-24

  articleSenior author

  This study introduces a novel iterative updating strategy powered by convolutional neural networks (CNNs) within a digital twin framework for informing structural behavior of infrastructure assets. A two-dimensional (2D) cantilever plate is employed as a benchmark case study for the updating strategy within a digital twin framework. Two distinct surrogate models are the main components of the digital twin framework, which form the basis of this work. The first model is formulated as finite element analysis surrogate, and the training dataset is prepared based on traditional finite element analysis. The second model predicts 2D deformation based on real-world surface images, serving as the ground truth, which is trained on pairs of deformed and reference images of surfaces with speckle patterns. This ground truth deformation data bridges the gap between experimental observations (physical twin) and virtual modeling (digital twin). The core novelty of this study lies in the iterative updating process and the introduction of four specialized CNNs, referred to as “calculators,” designed to refine the inputs for the first surrogate model using the ground truth displacement from the second surrogate model. Each calculator addresses a specific task: 1. Predicting loading in the y direction, given geometry, boundary conditions, loading in the x direction, and ground truth displacement. 2. Predicting loading in the x direction, given geometry, boundary conditions, loading in the y direction, and ground truth displacement. 3. Predicting geometry, given boundary conditions, loading, and ground truth displacement. 4. Predicting boundary conditions, given geometry, loading, and ground truth displacement. These calculators operate simultaneously in an iterative framework, using initial estimates of geometry, boundary conditions, and loading as inputs. Their outputs refine the inputs for subsequent iterations, with the process continuing until convergence. The final refined inputs are used by the first model, and their predictions are compared with the ground truth displacement from the second surrogate model to validate the digital twin. The results demonstrate that this novel approach integrating specialized calculators in an iterative updating process contributes to the robustness and accuracy of digital twins. The proposed updating strategy can pave the way for more reliable structural analysis and predictive modeling through digital twin frameworks, inform efficient infrastructure operation and management, and further contribute to the application and development of digital twin and structural health monitoring techniques.

  PublisherDOI

• Managing the Noise: Harnessing the Power of Graph Neural Networks to Detect Subsurface Damage Under Noisy Experimental Settings

  SSRN Electronic Journal · 2025-01-01

  preprintOpen accessSenior author
  PublisherDOI

• A Case Study on Leveraging Augmented Reality for Visualization in Structural Design

  2024-02-06 · 3 citations

  articleOpen accessSenior author

  She is currently working on building a digital twin that utilizes

  PublisherOA PDFDOI

• Deep Learning-Based Visual Measurement Extraction within an Adaptive Digital Twin Framework from Limited Data Using Transfer Learning

  SSRN Electronic Journal · 2024-01-01 · 1 citations

  preprintOpen access
  PublisherDOI

• Leveraging Mixed Reality for Augmented Structural Mechanics Education

  2021 ASEE Virtual Annual Conference Content Access Proceedings · 2024-02-20 · 6 citations

  articleOpen access

  Zijia graduated from the University of Virginia with a bachelor's degree in computer science

  PublisherOA PDFDOI

• Developing Augmented Reality Applications to Help Engineering Students Learn Spatial Structural Engineering Concepts

  2024-08-04

  articleOpen access

  In traditional mechanics-oriented classes, experience and the literature have shown that students are often challenged with conceptualizing complex three-dimensional behavior.Within the context of structural engineering and mechanics, the challenges manifest in scenarios related to linking this three-dimensional behavior with member response such as elastic buckling of columns and critical locations for shear and moment.While solutions such as props and videos have been used as examples in the past with some success, these tools do not spatially represent complex structural behaviors and are also limited to one-way interaction where the learner receives the information but cannot interact with the tools.This project leverages mobile augmented reality (AR) designed to help students visualize complex behaviors (deformation, strain, and stress) structural components with various loading and boundary conditions.The tool, STRUCT-AR utilizes finite element models pre-loaded into a mobile AR application that allows users to interact and engage with the models on their mobile device or tablet.Our vision of this technology is to provide a complementary teaching tool for enhancing personalized learning wherein students can leverage the technology as a learning companion both within the classroom and outside to better understand structural behaviors and mechanisms that are challenging to convey in a traditional 2D learning environment.This study uses a pilot study to evaluate how undergraduate and graduate students who have previously taken an introductory course on structural system design perceived the app.The purpose of this pilot study is to evaluate the usability of the app, its ability to improve spatial visualization ability, and to collect feedback on the app functionality.Study participants were asked to complete a pre and post-survey and the IBM Post-Study System Usability Questionnaire after engaging with the AR app on an iOS tablet.Results discuss how participants viewed the app in terms of its usability and usefulness and recommendations for tool refinement.Future work will be focused on conducting another pilot study after tool refinement before app deployment in a classroom setting.

  PublisherOA PDFDOI

• The impact of physician leadership development on behaviour and work-related changes

  Healthcare Management Forum · 2023-06-30 · 1 citations

  articleSenior author

  In this article, we present findings from a retrospective survey of 117 physician leadership development program graduates at the Sauder School of Business at the University of British Columbia in Vancouver. The survey was designed to assess how the program contributed to graduates' leadership development, specifically in terms of behaviour change and work-related changes. The themes resulting from the analysis of the open-ended questions reflected that the program led to changes in graduates' leadership behaviour and their ability to lead change in their respective organizations. The study highlighted the benefits of investment in training for physician leaders to advance transformation and improvement initiatives in a changing world.

  PublisherDOI

• Multi-modal deep fusion for bridge condition assessment

  Journal of Infrastructure Intelligence and Resilience · 2023-10-02 · 10 citations

  articleOpen access

  Bridge condition rating is a challenging task as it largely depends on the experience-level of the manual inspection and therefore is prone to human errors. The inspection report often consists of a collection of images and sequences of sentences (text) explaining the condition of the considered bridge. In a routine manual bridge inspection, an inspector collects a set of images and textual descriptions of bridge components and assigns an overall condition rating (ranging between 0 and 9) based on the collected information. Unfortunately, this method of bridge inspection has been shown to yield inconsistent condition ratings that correlate with inspector experience. To improve the consistency among image-text inspection data and further predict the accordant condition ratings, this study first provides a collective image-text dataset, extracted from the collection of bridge inspection reports from the Virginia Department of Transportation. Using this dataset, we have developed novel deep learning-base methods for an automatic bridge condition rating prediction based on data fusion between the textual and visual data from the collected report sets. Our proposed multi modal deep fusion approach constructs visual and textual representations for images and sentences separately using appropriate encoding functions, and then fuses representations of images and text to enhance the multi-modal prediction performance of the assigned condition ratings. Moreover, we study interpretations of the deployed deep models using saliency maps to identify parts of the image-text inputs that are essential in condition rating predictions. The findings of this study point to potential improvements by leveraging consistent image-text inspection data collection as well as leveraging the proposed deep fusion model to improve the bridge condition prediction rating from both visual and textual reports.

  PublisherDOI

Frequent coauthors

• Mohamad Alipour

  Tongji University

  86 shared

• Mehrdad Shafiei Dizaji

  University of Massachusetts Lowell

  76 shared

• Ayatollah Yehia

  University of British Columbia

  52 shared

• Osman E. Ozbulut

  39 shared

• Amir Gheitasi

  36 shared

• Connor Lyons

  Engineering Systems (United States)

  36 shared

• Jacqueline Chao

  University of Tabriz

  36 shared

• Theresa M. Ahlborn

  Michigan Technological University

  23 shared

Awards & honors

• Delmar L. Bloem Distinguished Service Award 2021
• IAspire Leadership Academy Fellow 2020–2022
• Outstanding Reviewer - American Society of Civil Engineering…
• Excellence in Diversity Fellowship - University of Virginia…
• ACI Young Member Award for Professional Achievement - Americ…