Adaptively Informed Trees (AIT*) and Effort Informed Trees (EIT*): Asymmetric bidirectional sampling-based path planning

Optimal path planning is the problem of finding a valid sequence states between start and goal that optimizes an objective. Informed algorithms order their search with problem-specific knowledge expressed as heuristics can be orders magnitude more efficient than uninformed algorithms. Heuristics are most effective when they both accurate computationally inexpensive to evaluate, but these often conflicting characteristics. This makes selection appropriate difficult for many problems. paper presents two almost-surely asymptotically optimal sampling-based address this challenge, Adaptively Trees (AIT*) Effort (EIT*). These use asymmetric bidirectional in which searches continuously inform each other. allows AIT* EIT* improve performance by simultaneously calculating exploiting increasingly accurate, heuristics. The benefits relative other demonstrated on twelve problems abstract, robotic, biomedical domains optimizing length obstacle clearance. experiments show outperform clearance, where priori cost ineffective, still perform well minimizing length, such effective.