All Issue

2025 Vol.4, Issue 4

Research Article

BIM-Point Cloud Fusion-based Automated Training Dataset Construction and Segmentation-based Weld Recognition

BIM-포인트 클라우드 융합 기반 학습 데이터셋 구축 자동화 및 세그멘테이션 기반 용접부 인식

Seungho Choi1
,
Seungjun Ahn2*
최 승호1
,
안 승준2*
1Bachelor’s Student, Department of Civil and Environmental Engineering, Hongik University
2Corresponding Author, Member, Associate Professor, Department of Civil and Environmental Engineering, Hongik University
1홍익대학교 건설환경공학과 학사과정
2교신저자 ‧ 정회원 ‧ 홍익대학교 건설환경공학과 부교수
*Corresponding Author
30 December 2025. pp. 1-8
PDF XML Citation
Abstract

To address key challenges in construction sites, including ensuring safety in high-elevation welding operations and mitigating the shortage of skilled welders, a foundational technology for robot-based welding automation is proposed. An automated pipeline for constructing training datasets is developed by integrating Building Information Modeling (BIM) and three-dimensional point cloud technologies, and a semantic segmentation–based 3D weld recognition model is designed and applied to accurately extract the location and geometry of welds within steel connections. Using BIM design information, point cloud datasets of steel connections including weld regions are systematically generated, and the trained segmentation model identifies the geometric shape and spatial location of welds with high accuracy. By linking BIM and artificial intelligence technologies, this research establishes a foundation for construction automation and provides a core technical basis for advancing task perception in intelligent welding robots and enhancing the fidelity of digital twin implementations in the construction industry.

Keywords
High-elevation welding automation
Semantic segmentation
BIM
Point Cloud
Steel connection
Weld recognition

본 연구는 건설 현장의 주요 과제인 고소 용접 작업의 안전성 확보와 숙련공 부족 문제를 해결하기 위해, 로봇 기반 용접 자동화를 위한 기반 기술을 제안한다. BIM 모델링과 3차원 포인트 클라우드 기술을 융합한 학습 데이터셋 자동 구축 파이프라인을 제시하고, 시멘틱 세그멘테이션 기반의 3차원 용접부 인식 모델을 구성 및 적용하여 철골 접합부 내 용접부의 위치와 형상을 정밀하게 추출한다. BIM의 설계 정보를 활용해 용접부를 포함한 철골 접합부의 포인트 클라우드 데이터셋을 체계적으로 생성하며, 세그멘테이션 모델은 이를 학습하여 용접부의 기하학적 형상과 위치를 높은 정확도로 식별한다. 본 연구는 BIM과 인공지능 기술을 연계함으로써 건설 자동화의 기반을 마련한 연구로, 향후 지능형 용접 로봇의 작업 인지 기술 및 건설 분야 디지털 트윈 구현의 완성도를 높이는 핵심 기술적 토대를 제공한다.

키워드
고소 용접 자동화
시멘틱 세그멘테이션
BIM
포인트 클라우드
철골 접합부
용접부 인식
References
1

Ahmed, S. M., Chew, C. M., Tan, Y. Z., and Mamun, A. A. (2018). Edge and corner detection for robotic welding. Proceedings of the IEEE International Conference on Robotics and Automation, pp. 187-194.

2

Benallal, H., Saab, N. A., Tairi, H., Alfalou, A., and Riffi, J. (2025). Advancements in semantic segmentation of 3D point clouds for scene understanding using deep learning. Technologies, 13(8), 322. https://doi.org/10.3390/technologies13080322

10.3390/technologies13080322
3

Chang, A. X., Funkhouser, T., Guibas, L., Hanrahan, P., Huang, Q., Li, Z., Savarese, S., Savva, M., Song, S., Su, H., Xiao, J., Yi, L., and Yu, F. (2015). ShapeNet: An information-rich 3D model repository. arXiv preprint arXiv:1512.03012.

4

Li, T., Meng, S., Lu, C., Wu, Y., and Liu, J. (2025). A novel BIM and vision-based robotic welding trajectory planning method for complex intersection curves. Measurement, 253, 117587. https://doi.org/10.1016/j.measurement.2025.117587

10.1016/j.measurement.2025.117587
5

Lin, T.-Y., Goyal, P., Girshick, R., He, K., and Dollár, P. (2017). Focal loss for dense object detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(2), pp. 318-327. https://doi.org/10.1109/ICCV.2017.324

10.1109/ICCV.2017.324
6

Ministry of Employment and Labor. (2025). Industrial accident statistics as of the end of June 2025: Supplementary statistics on fatal occupational accidents subject to investigation (provisional). Ministry of Employment and Labor.

7

Mo, K., Zhu, S., Chang, A. X., Yi, L., Tripathi, S., Guibas, L. J., and Su, H. (2019). PartNet: A large-scale benchmark for fine-grained and hierarchical part-level 3D object understanding. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 909-918. https://doi.org/10.1109/CVPR.2019.00100

10.1109/CVPR.2019.00100
8

Oh, B., Kim, M., Lee, C., Cho, H., and Kang, K.-I. (2019). 3D modeling approach of building construction based on point cloud data using LiDAR. Proceedings of the 36th International Symposium on Automation and Robotics in Construction (ISARC 2019), Tokyo, Japan, pp. 906-913. https://doi.org/10.22260/ISARC2019/0121

10.22260/ISARC2019/0121
9

Qi, C. R., Su, H., Mo, K., and Guibas, L. J. (2017a). PointNet: Deep learning on point sets for 3D classification and segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 652-660.

10

Qi, C. R., Yi, L., Su, H., and Guibas, L. J. (2017b). PointNet++: Deep hierarchical feature learning on point sets in a metric space. Advances in Neural Information Processing Systems, 30, pp. 5099- 5108.

11

Sudre, C. H., Li, W., Vercauteren, T., Ourselin, S., and Cardoso, M. J. (2017). Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations. Proceedings of the International Workshop on Deep Learning in Medical Image Analysis, pp. 240-248. https://doi.org/10.1007/978-3-319-67558-9_28

10.1007/978-3-319-67558-9_2834104926PMC7610921
12

Sung, Y.-K., Park, H.-D., and Choi, S.-Y. (2025). Trends in changes of technical manpower in construction sites and strategies for securing workforce. Construction & Economy Research Institute of Korea.

13

Takubo, T., Miyake, E., Ueno, A., and Kubo, M. (2023). Welding line detection using point clouds from optimal shooting position. Journal of Robotics and Mechatronics, 35(2), pp. 492-500. https://doi.org/10.20965/jrm.2023.p0492

10.20965/jrm.2023.p0492
14

Wang, Y., Sun, Y., Liu, Z., Sarma, S. E., Bronstein, M. M., and Solomon, J. M. (2019). Dynamic graph CNN for learning on point clouds. ACM Transactions on Graphics, 38(1), Article 1. https://doi.org/10.1145/3326362

10.1145/3326362
15

Zhou, N., Jiang, P., Jiang, S., Shu, L., Ni, X., and Zhong, L. (2024). An identification and localization method for 3D workpiece welds based on the DBSCAN point cloud clustering algorithm. Journal of Manufacturing and Materials Processing, 8(6), 287. https://doi.org/10.3390/jmmp8060287

10.3390/jmmp8060287
Information
  • Publisher :Korean Society of Automation and Robotics in Construction
  • Publisher(Ko) :(사)한국건설자동화·로보틱스학회
  • Journal Title :Journal of Construction Automation and Robotics
  • Journal Title(Ko) :건설자동화·로보틱스 논문집
  • Volume : 4
  • No :4
  • Pages :1-8
  • Received Date : 2025-12-23
  • Revised Date : 2025-12-29
  • Accepted Date : 2025-12-29
  • DOI :https://doi.org/10.55785/JCAR.4.4.1
Journal Informaiton Journal of Construction Automation and Robotics Journal of Construction Automation and Robotics
Online Submission
  • crossref crossmark
  • crossref cited-by
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close