Answers are grounded in this paper and cite their sources.
Automated assembly of complex block structures is important for improving construction efficiency, precision, and long-duration autonomous operation in unstructured environments. However, reliable assembly remains challenging because discrete irregular blocks with random positions and orientations are simultaneously constrained by sequence dependencies, temporary support requirements, and evolving collision boundaries during construction. To address these challenges, this study proposes a vision-guided task-level robotic assembly framework for automated construction of complex block structures. The framework integrates hierarchical visual perception for coarse-to-fine component recognition and pose estimation, constraint-aware assembly planning for topology-guided sequence generation and collision-free asymptotically optimal path planning, and vision-guided pose compensation strategy during execution. Experiments on three representative structures achieved mAP50 above 99.0%, IoU above 98.0%, 100% sequence-generation success, planning efficiency comparable to joint-space RRT*, generally shorter paths, and millimeter-level placement accuracy during assembly, demonstrating the feasibility and practical value of the proposed framework for automated construction of componentized building systems.
No takes yet. Share an insight, caveat, or question.
Zhu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: