About

Professor Nora El-Gohary earned her B.Sc. and M.Sc. degrees from the American University in Cairo in Construction Engineering, in 1999 and 2002 respectively. She completed her Ph.D. in Civil Engineering at the University of Toronto in 2008. She joined the Department of Civil and Environmental Engineering at the University of Illinois at Urbana-Champaign in December 2009, after serving as an Assistant Professor at the University of Manitoba. Her professional experience includes working for Hyundai Engineering & Construction Co. Ltd, one of the largest construction and civil engineering companies. Her research interests focus on data analytics and artificial intelligence (AI) for supporting sustainable and value-adding civil infrastructure systems. Her work encompasses building information modeling, semantic data and information modeling, natural language processing, machine learning, data integration, and human-centered systems and analytics. She has contributed to the development of automated construction management systems and has been recognized with numerous research awards, including the NSF CAREER Award and the Dean's Award for Excellence in Research. Dr. El-Gohary has secured research sponsorship from agencies such as the NSF, Illinois Department of Transportation, Qatar Foundation, and the Natural Sciences and Engineering Research Council of Canada. As a member of the Construction Engineering and Management faculty, she teaches courses related to construction cost analysis and data modeling. She has held leadership roles in professional societies, including serving as Chair of the Executive Committee of the ASCE's Computing Division and as an Associate Editor of the Journal of Computing in Civil Engineering. Her academic positions include Assistant Professor, Associate Professor, and Professor at the University of Illinois, with prior experience at the University of Manitoba. She is a licensed Professional Engineer in Ontario and Egypt.

Research topics

• Computer Science
• Machine Learning
• Artificial Intelligence
• Data Mining
• Engineering
• Environmental science
• Physics
• Automotive engineering
• Thermodynamics
• Mathematics
• Statistics
• Database
• Electrical engineering

Selected publications

• Toward open science in the AEC community: An ecosystem for sustainable digital assets sharing and reuse

  Developments in the Built Environment · 2026-03-19

  articleOpen access

  The Architecture, Engineering, and Construction (AEC) industry is undergoing rapid digital transformation, generating diverse digital assets such as datasets, computational models, workflows, and educational resources. However, these resources are often fragmented across repositories and inconsistently documented, limiting their discoverability, interpretation, and reuse in research, education, and practice. This study introduces OpenConstruction , a community-driven open science ecosystem that aggregates and organizes openly accessible AEC digital assets. The platform integrates four catalogs, including datasets, models, workflows, and educational resources, supported by standardized metadata, curator-led validation and community-based governance processes that ensure transparency and resource quality. As of December 2025, the ecosystem indexes more than 200 digital assets spanning research, education, and industry applications. Two case studies illustrate how the platform supports resource discovery, benchmarking, and instructional use. The platform is publicly accessible at https://www.openconstruction.org/ . • Open ecosystem for AEC datasets, models, workflows, and OERs. • Standardized catalogs improve discovery, comparability, and reuse in AEC. • Case studies show support for benchmarking and curriculum development.

  PublisherDOI

• Automated free-form bridge inspection image interpretation using adaptive CNN, transformer feature fusion, and contrastive learning

  Automation in Construction · 2026-03-19

  articleSenior authorCorresponding
  PublisherDOI

• Toward Open Science in the AEC Community: An Ecosystem for Sustainable Digital Knowledge Sharing and Reuse

  arXiv (Cornell University) · 2026-01-02

  preprintOpen access

  The Architecture, Engineering, and Construction (AEC) industry is undergoing rapid digital transformation, producing diverse digital assets such as datasets, computational models, use cases, and educational materials across the built environment lifecycle. However, these resources are often fragmented across repositories and inconsistently documented, limiting their discoverability, interpretability, and reuse in research, education, and practice. This study introduces OpenConstruction, a community-driven open-science ecosystem that aggregates, organizes, and contextualizes openly accessible AEC digital resources. The ecosystem is structured into four catalogs, including datasets, models, use cases, and educational resources, supported by consistent descriptors, curator-led validation, and transparent governance. As of December 2025, the platform hosts 94 datasets, 65 models, and a growing collection of use cases and educational materials. Two case studies demonstrate how the ecosystem supports benchmarking, curriculum development, and broader adoption of open-science practices in the AEC sector. The platform is publicly accessible at https://www.openconstruction.org/.

  PublisherDOI

• Toward Open Science in the AEC Community: An Ecosystem for Sustainable Digital Knowledge Sharing and Reuse

  ArXiv.org · 2026-01-02

  articleOpen access

  The Architecture, Engineering, and Construction (AEC) industry is undergoing rapid digital transformation, producing diverse digital assets such as datasets, computational models, use cases, and educational materials across the built environment lifecycle. However, these resources are often fragmented across repositories and inconsistently documented, limiting their discoverability, interpretability, and reuse in research, education, and practice. This study introduces OpenConstruction, a community-driven open-science ecosystem that aggregates, organizes, and contextualizes openly accessible AEC digital resources. The ecosystem is structured into four catalogs, including datasets, models, use cases, and educational resources, supported by consistent descriptors, curator-led validation, and transparent governance. As of December 2025, the platform hosts 94 datasets, 65 models, and a growing collection of use cases and educational materials. Two case studies demonstrate how the ecosystem supports benchmarking, curriculum development, and broader adoption of open-science practices in the AEC sector. The platform is publicly accessible at https://www.openconstruction.org/.

  PublisherOA PDF

• Automated Window Opening Segmentation and Levelness Measurement

  2026-01-28

  articleSenior author

  Conventional wall opening inspections, including levelness checks, rely on manual surveying, which is time-consuming and labor-intensive. Recent research has explored automated image-based methods, with image segmentation playing a key role. Red-green-blue-depth (RGB-D) images, which include depth information, have been used to improve segmentation accuracy. However, these efforts are limited in using depth maps in measurement tasks. To address this gap, this paper proposes a deep learning-based method for detecting wall openings and measuring their levelness using RGB-D images. The approach includes (1) segmenting window openings in indoor scenes using a deep learning-based model, and (2) identifying window corners and computing levelness based on depth values and a horizontal baseline. The method was evaluated on real-world indoor scenes, achieving 85.7% intersection over union (IoU) for segmentation, 70.52 mm mean absolute error (MAE), and 3.52% mean relative error (MRE) for measurement.

  PublisherDOI

• Personalized thermal comfort prediction using occupant features from video recordings

  Building and Environment · 2026-02-02

  articleSenior author
  PublisherDOI

• Deep Learning-Based Unsupervised Domain-Adaptive Method for Corrosion Segmentation from Real-World Images

  2026-01-28

  articleSenior author

  Corrosion poses a significant threat to infrastructure, compromising its safety and longevity. Traditional corrosion assessment methods, which rely on subjective visual inspections and manual measurements, are time-consuming and inefficient. Advancements in artificial intelligence (AI) have demonstrated potential for automated and accurate corrosion detection and analysis. However, the development of deep learning-based methods for corrosion segmentation has several challenges. First, corrosion areas are irregular and often blend into other damages, making them difficult to segment. Second, supervised learning approaches require extensive pixel-level labeled datasets, which involve labor-intensive annotation processes. To address these limitations, this paper proposes a deep learning-based unsupervised domain-adaptation framework for automatic corrosion segmentation from real-world images. The proposed approach employs deep kernel maximum mean discrepancy for feature alignment. The experimental results demonstrated promising performance, suggesting the framework’s potential to reduce the reliance on labeled data and to effectively address domain discrepancies for accurate corrosion segmentation.

  PublisherDOI

• Introducing Data Papers: A New Article Type in the <i>Journal of Computing in Civil Engineering</i>

  Journal of Computing in Civil Engineering · 2026-02-04

  articleOpen accessSenior author
  PublisherDOI

• Text-Enhanced Label-Efficient Automated Bridge Defect Semantic Segmentation from Inspection Images

  2025-12-11

  articleSenior authorCorresponding

  The utilization and integration of unmanned aerial vehicles (UAVs) and computer vision technologies in recent automated bridge inspection methodologies has shown advancement in capturing and analyzing images to enhance the efficiency and safety of bridge inspection. However, information extraction from inspection images collected on-site remains challenging. First, although extensive research efforts have focused on segmenting defects from images, the localization and segmentation performance is limited due to complex backgrounds and irregular defect shapes in images. Second, precise pixel-level annotation of defect masks is labor-intensive and time-consuming, which underscores the need for a label-efficient method for defect segmentation. To address these gaps, this paper proposes a deep learning-based method to extract and segment different types of bridge defects from on-site inspection images using a label-efficient way, which leverages corresponding text descriptions, the Grounding DINO (DETR with Improved deNoising anchOr boxes) object detection model, and the segment anything model (SAM). This paper discusses the proposed method and its performance results. The experimental results show that the method can efficiently extract and segment various bridge defects, which would support automated bridge inspection.

  PublisherDOI

• Generative Model-Based Data Augmentation Approach for Supporting Maintenance Decision-Making

  2025-12-11

  articleSenior authorCorresponding

  Textual bridge reports encompass important and detailed technical data and information about bridge conditions and maintenance activities. These reports not only serve as critical sources of information and insights but also present an opportunity for improving the understanding of bridge deterioration and maintenance, and hence, for improving the decision-making support within the domain of bridge maintenance. However, obtaining a sufficiently large and diverse data set of reports/text that would capture different maintenance scenarios and contexts is often challenging due to the limited availability of real-world data. To better support maintenance decision-making, there is a need for data augmentation to address the challenges associated with limited training data. To address this need, this paper proposes a generative model-based data augmentation approach to enhance the size and quality of training data sets for supporting the development of deep learning models. The generative model draws new cases/scenarios from existing domain knowledge such as bridge inspection and maintenance guidance documents and specifications. This paper discusses the proposed approach and the experimental results, including the evaluation of the generated synthetic data.

  PublisherDOI

Recent grants

• Convergence Accelerator Phase I (RAISE): Civil Infrastructure Systems Open Knowledge Network (CIS-OKN)

  NSF · $1000k · 2019–2021

• Deontic Modeling and Natural Language Processing for Automated Environmental and Green Compliance Checking

  NSF · $320k · 2012–2016

• CAREER: Axiological Modeling and Simulation for Value-Sensitive Infrastructure Project Planning and Design

  NSF · $400k · 2013–2019

Frequent coauthors

• Jiansong Zhang

  46 shared

• Kaijian Liu

  Stevens Institute of Technology

  35 shared

• Mani Golparvar Fard

  University of Illinois Urbana-Champaign

  26 shared

• Pingbo Tang

  Carnegie Mellon University

  22 shared

• Mario Bergés

  Carnegie Mellon University

  21 shared

• Hubo Cai

  China Electric Power Research Institute

  21 shared

• Jiannan Cai

  The University of Texas at San Antonio

  21 shared

• Yanyu Wang

  21 shared

Awards & honors

• Natural Sciences and Engineering Research Council of Canada…
• National Science Foundation CAREER Award (2013)
• Center of Advanced Study Fellow (2015)
• Excellence Faculty Fellow in Civil and Environmental Enginee…
• Campus Distinguished Promotion Award (2017)