Proceedings of the 2021 European Conference on Computing in Construction


Determining construction method patterns to automate and optimise scheduling – a graph-based approach

Ying Hong 1, Vahan Hovhannisyan 2, Haiyan Xie 3, Ioannis Brilakis 1
1 University of Cambridge, United Kingdom
2 nPlan, United Kingdom
3 Illinois State University, Illinois, United States

DOI: 10.35490/EC3.2021.165
Abstract: Creating quality construction schedules to mitigate delays often relies on experience. The lack of dissemination of historic decision reasoning makes it harder. This study proposes a graph-based method to find the time- and risk-efficient construction method patterns from historic projects to help schedulers improve productivity and accuracy. The method leverages schedule data obtained from a Tier-1 contractor and validates for excavation activities. The results indicate that the most time-efficient excavation activities can be done in 0.6% of total project time. The proposed method can help industry professionals standardise scheduling guidelines and automate the generation of construction schedules for critical subtasks.
Keywords: Scan-to-BIM, TLS, point cloud, semantic segmentation, synthetic data
Pages: 59 - 66

Presentation Video: