Answers are grounded in this paper and cite their sources.
Time-critical unmanned aerial vehicle missions often require a vehicle to reach a specified region within a prescribed arrival-time window, rather than only finding a collision-free path. Under a fixed-speed flight assumption, this requirement makes arrival-time feasibility directly dependent on path geometry. This paper proposes a time-window-aware rapidly exploring random tree star (TWA-RRT*) algorithm. The proposed planner embeds the target arrival-time window into tree construction using a look-ahead time-window-aware parent selection criterion, cost-consistent rewiring with recursive propagation of path-dependent quantities, late-infeasible pruning based on an arrival-time lower bound, and early-arrival loitering near the goal region. These mechanisms guide sampling-based exploration toward branches that are spatially collision-free and temporally feasible, while eliminating branches that cannot satisfy the latest allowable arrival time. The method was evaluated in four benchmark cases, comprising moderate, narrow-passage, and cluttered obstacle configurations and an additional early-arrival case, over 100 randomized trials per case. The proposed method achieved success rates of 100%, 94%, 86%, and 100%, respectively, with average computation times of 1.346, 1.984, 2.226, and 5.680 s. In the early-arrival case, TWA-RRT* achieved a 100% success rate with zero time-window violations, demonstrating the effectiveness of the proposed loitering mechanism. Compared with representative temporal planning baselines, the proposed method provided a favorable balance between computational efficiency and time-window feasibility. The results demonstrate the effectiveness of incorporating arrival-time-window awareness directly into sampling-based unmanned aerial vehicle path planning.
No takes yet. Share an insight, caveat, or question.
Taegeun Oh (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: