Target-Focused Local Workspace Modeling for Construction Automation Applications

Abstract

To improve the efficiency and effectiveness of the geometric data collection and processing for automated construction operations, a light-weight hybrid light amplification for detection and ranging (LADAR) system was developed to enable semi-automatic targeted data collection, such that an operator can quickly obtain high-quality data for parts of the workspace critical for an on-going operation. To further improve the precision of separating needed data from background and irrelevant data, using this custom-designed system, the relationships were investigated among the amplitude of data points, the scanning distance, time-of-flight, and the grayscales of objects. Thus, positive control of scanning position, orientation, field of view and amplitude is enabled. For construction robotics applications in confined spaces such as power plants, such positive control is a critical capability. The evaluation results indicate the efficiency and effectiveness of the semi-automatic targeted scanning approach on reducing the scanning time and data size, and the effectiveness of the proposed target-focused data collection approach.