Facilitating the Information Exchange Between A BIM and Construction Robot
Restricted (Penn State Only)
- Author:
- Mc Clymonds, Austin
- Graduate Program:
- Architectural Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- November 15, 2023
- Committee Members:
- James Freihaut, Program Head/Chair
Jose Pinto Duarte, Outside Unit & Field Member
Rob Leicht, Major Field Member
Somayeh Asadi, Chair & Dissertation Advisor
John Messner, Major Field Member - Keywords:
- Building Information Modeling (BIM)
Robots
Information Exchange
Parametric Modeling
Computational Modeling
Masonry
Construction
Robotic Construction
System Architecture
Photogrammetry
Reality Capture - Abstract:
- Historically, the construction industry has lagged behind other sectors in technology adoption. In recent years, a transformative shift, often referred to as Construction 4.0, has taken place, bringing automation and digitization, including robotics, into the industry. Despite these advancements, integrating robots effectively into construction sites remains a challenge. One fundamental issue is how robots receive and interpret information, especially given their lack of inherent construction knowledge. This research addresses the critical challenge of enabling robots to comprehend and execute tasks based on construction-related data. To this end, the primary objective of this study is to develop a method for transferring data from the Building Information Model (BIM) to robots, thereby facilitating task execution with a specific focus on the model's Level of Development (LOD). A systematic literature was conducted as the first step to perform this research, which explored existing system architectures related to BIM and robotics. It was necessary to review both types of system architecture to ensure interoperability across information platforms and to facilitate robotic task execution. The system architecture was developed by integrating facets from both domains and leveraging existing knowledge within the construction industry. This architecture encompassed three key steps: task planning, task decomposition, and robotic task execution, with two crucial information exchanges bridging the key steps. From there, information exchanges were further developed to support the process. Specifically, this research sought to increase the LOD of a masonry wall from level 200 to 400, which contains information related to the construction process and eventually facilitates robotic task execution. This allowed for extracting important model information such as location, rotation, element identification (I.D), and element type. Computational tools, particularly the computational modeler Dynamo, were instrumental in developing a script to increase LOD. Using the developed script, the LOD of the 3D masonry model developed in Autodesk Revit was enhanced to LOD 400 for multiple masonry wall configurations across two types of masonry units. Notably, the research highlighted that computational processing time increased proportionally as model complexity rose. In the final phase of the study, real-world experiments were conducted using a Clearpath Husky A200 robot. The investigation focused on parameters including distance from the wall, image spacing, robot path, and wall configuration to identify viable methods for collecting photometric data. Each wall configuration was designed in Autodesk Revit, where information was extracted from the model. The location data from the model facilitated on-site wall construction based on the local coordinate system established in the model. The Husky robot was teleoperated to collect images and Global Positioning Systems (GPS) data, which was utilized to develop point clouds for each configuration. Subsequently, data from each point cloud was extracted and filtered to determine optimal data collection methods. This study explores the challenges and solutions of integrating robots into construction processes specific to information requirements. By developing a robust information exchange method and conducting real-world experiments, this research contributes valuable insights to the growing field of Construction 4.0 and advances the understanding of effective robotic task execution on the construction site.