Autonomous productivity measurement NDT-SLAM-based algorithm for excavator earthwork operations

Earthwork operations are essential components of construction projects, with real-time monitoring of earthwork productivity being crucial for effective project control and resource management. Excavators play a central role in excavation and earthmoving activities. Although several advanced monitoring methods have been introduced, they require post-processing, limiting their ability to provide real-time productivity data.. This paper presents an NDT-SLAM-based algorithm designed for real-time productivity monitoring of excavator earthwork operations. The method generates real-time point clouds of the earthwork environment using a single RGB-D camera, allowing continuous tracking of excavation progress by recording the displacement of earthen material over time. It compares changes in volume over time to estimate productivity. To evaluate the proposed method's accuracy and efficiency, it was implemented in a test bed environment and compared with laser scanning as ground truth. The findings indicate that the proposed method reduced real-time data collection time by approximately 45%, while maintaining volume and productivity measurements within acceptable error margins for field applications. Furthermore, to validate its applicability in an earthwork environment, an implementation case study on an earthwork project is presented. The results indicate that the proposed method efficiently calculated the volume and estimated real-time productivity.