ASIP-Planner lets UAVs adapt to unknown obstacles in indoor inspections

Manual inspection of indoor infrastructure like tunnels and industrial facilities is costly and dangerous. Drones offer an alternative, but they typically rely on floorplan-derived maps that fail to account for temporary obstacles such as equipment or scaffolding. Existing coverage planners assume a fully known environment and perform deterministic global optimization, making them brittle when real conditions differ from the map.

ASIP-Planner solves this by integrating two complementary modules. The global planner segments planar inspection targets into surface-aligned clusters, generating compact viewpoint sequences with improved orientation consistency. The local planner then generates collision-free trajectories and adjusts the viewing angle online to mitigate occlusion-induced coverage loss, while preserving the planned trajectory structure. In simulations across randomized scenes, the approach achieved near-complete coverage with shorter flight paths than baselines. Real-world flight experiments confirmed the framework produces high-quality inspection data for downstream analysis.