Autonomous Robotics & Computer Vision in Construction: Simultaneous Localization and Mapping (SLAM), Photogrammetric Bundle Adjustment, and Robotic End-Effector Trajectory Kinetics
Autonomous robotics and computer vision in construction engineering leverage spatial perception algorithms, real-time sensor fusion, and automated execution systems to transform jobsite monitoring, structural inspection, Earthwork operations, and digital-twin verification. By integrating LiDAR-driven spatial mapping, unmanned aerial vehicles (UAVs), legged quadruped robots, and robotic manipulators, autonomous systems enable continuous site progress tracking and automated structural assembly while minimizing human exposure to hazardous site conditions. Autonomous navigation and spatial modeling across unstructured construction environments rely on Visual-Inertial Simultaneous Localization and Mapping (VI-SLAM) . The robot pose state vector $\mathbf{x}_k$ at time step $k$ is optimized alongside 3D landmark points $\mathbf{p}_j$ by minimizing the non-linear reprojection error $\mathbf{e}_{ij}$ across camera frames using Bundle Adjustment : $$\min_{\mathbf{x}, \mathbf{p}} \sum_{i} \s...