AbstractBuilding Information Modeling (BIM) has greatly aided lean management
in construction, providing cost savings, efficiency improvements, collaboration, and sustainability throughout the building lifecycle. Several countries and industries promote the development and practical application of
BIM. However, the current scan-based reconstruction process of BIM requires costly specialized equipment and extensive manual data collection
and processing, hindering rapid reconstruction and continuous updating.
As a solution, Autonomous Mobile Robots (AMR) are expected to be the
most promising platform for autonomous 3D BIM reconstruction.
This thesis achieves a brief review of robot-based BIM reconstruction
methods through a data flow-based classification approach, identifying three
main challenges to robots: limited Field of View (FOV) of sensors, lack
of scan quality assessment, and rough autonomous movement control. To
address these challenges, a novel rotating Light Detection and Ranging
(LiDaR) gimbal design is presented, along with scan distortion removal
and scan quality evaluation algorithms. A three-step autonomous navigation method that integrates scan quality and scan parameters is proposed,
enabling efficient, reconstruction-oriented autonomous navigation. A robot
is designed and built to validate the proposed algorithms, and real-world
testing demonstrates their effectiveness.
|Date of Award||Oct 2023|
|Supervisor||Adam Rushworth (Supervisor)|
- BIM reconstruction
- Mobile Robotic
- Optimal Scanning Parameters
- LiDAR Gimbal
- Autonomous Navigation