{"id":"3a149a0c-77be-48da-ac50-b58b3787633e","arxiv_id":"2608.00625","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A status map of 2015-2025 learning-based motion planning in dynamic environments, organized by four roles learning can play: direct policy, classical-planner augmentation, hybrid coupling, and training support.","lead":"This paper surveys a decade of research on robots planning paths and motions among moving people, vehicles, and other robots, and sorts the field into four roles that machine learning can play. It is a reference map for robotics researchers and engineers choosing a learning-plus-planning recipe for their platform.","discovery_kind":"review","skeptic_critique":null,"referee_report":null,"author_rebuttal":null,"desk_editor":null,"rs_alignment":null,"lean_confirmation":null,"pith_extraction":null,"created_at":"2026-08-04T01:47:44.112623+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":null,"supporting_citations":[],"review_version":1}