Doyun Lee

Dr. Doyun Lee

Civil Engineering and Construction, Georgia Southern University, Statesboro, Georgia, USA
orcid

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 September 2024 in Automation in Construction
Verified
Journal Article
Closed access
Doyun LeeDoyun Lee
Doyun Lee
KHKevin Han
Kevin Han


1 March 2024 in Journal of Construction Engineering and Management
Verified
Journal Article
Closed access
Doyun LeeDoyun Lee
Doyun Lee
GNGuang-Yu Nie
Guang-Yu Nie
Kevin HanKevin Han
Kevin Han

1 May 2025 in Journal of Construction Engineering and Management
Verified
Journal Article
Closed access
Doyun LeeDoyun Lee
Doyun Lee
Kevin HanKevin Han
Kevin Han

9 November 2020
Verified
Conference Paper
Closed access
PDF file
Doyun LeeDoyun Lee
Doyun Lee
KHKevin Han
Kevin Han
In modular construction, poor quality assurance and quality control (QA/QC) at offsite facilities can lead to serious cost escalations and delays. Offsite components that are not properly inspected will lead to incompatibility issues on job site during assemblies, which will lead to re-fabrication and re-shipping. This added lead time can easily affect the critical path. There have been research efforts to perform a visual inspection (although not specific to modular construction) using high precision sensors (i.e., laser scanners) and image-based 3D reconstruction. The former solution can be an expensive solution for smaller firms. The latter solution results in inconsistent results because it results in relatively noisy data. This paper attempts to address this limitation by combining leveraging camera poses from image-based 3D reconstruction and their images to select boundaries to be inspected. The proposed method is validated through a case study. A high precision mobile laser scanner was used for collecting the ground truth. The results demonstrate the effectiveness of the proposed method.

13 January 2026 in Sensors
Verified
Journal Article
Open access
PDF file
EOEmmanuella Ogun
Emmanuella Ogun
YVYong Ann Voeurn
Yong Ann Voeurn
Doyun LeeDoyun Lee
Doyun Lee
The deterioration of civil infrastructure poses a significant threat to public safety, yet conventional manual inspections remain subjective, labor-intensive, and constrained by accessibility. To address these challenges, this paper presents a real-time robotic inspection system that integrates deep learning perception and autonomous navigation. The proposed framework employs a two-stage neural network: a U-Net for initial segmentation followed by a Pix2Pix conditional generative adversarial network (GAN) that utilizes adversarial residual learning to refine boundary accuracy and suppress false positives. When deployed on an Unmanned Ground Vehicle (UGV) equipped with an RGB-D camera and LiDAR, this framework enables simultaneous automated crack detection and collision-free autonomous navigation. Evaluated on the CrackSeg9k dataset, the two-stage model achieved a mean Intersection over Union (mIoU) of 73.9 ± 0.6% and an F1-score of 76.4 ± 0.3%. Beyond benchmark testing, the robotic system was further validated through simulation, laboratory experiments, and real-world campus hallway tests, successfully detecting micro-cracks as narrow as 0.3 mm. Collectively, these results demonstrate the system’s potential for robust, autonomous, and field-deployable infrastructure inspection.