UniConFlow is a unified constrained flow‑matching framework that generates robot trajectories while systematically handling both equality and inequality constraints such as collision avoidance, actuation limits, and dynamic consistency. It introduces a prescribed‑time zeroing function to shape a time‑varying guidance field during inference, and employs a violation‑segment extraction protocol and trajectory compression to address computational challenges in long‑horizon, high‑dimensional planning. Experiments on a double inverted pendulum, a real‑to‑sim car racing task, and a sim‑to‑real manipulation task show that UniConFlow outperforms state‑of‑the‑art generative planners and conventional optimization baselines on certified motion planning metrics like safety, kinodynamic consistency, and action feasibility.
By Zewen Yang, Xiaobing Dai, Dian Yu, Zhijun Li, Majid Khadiv, Sandra Hirche, Sami Haddadin
Recent advances in generative planning have made trajectory inpainting a promising approach to offline goal-conditioned reinforcement learning. However, these methods typically specify the planning ho...
HorizonFlow is a hierarchical planner for offline goal-conditioned reinforcement learning that treats the planning horizon as an output rather than a fixed input. It uses a subgoal route planner and an action-prefix controller, both employing insertion-based generation and flow matching, to jointly generate continuous plan content and its length. The method leverages the partially generated plan to guide token insertion and to steer generation toward shorter plans, achieving superior performance on Maze2D, Multi2D, and OGBench benchmarks.
By JunHyeok Oh, Zian Jang, Byung-Jun Lee
arXiv:2504. 16738v3 Announce Type: replace-cross Abstract: Planning long-horizon manipulation motions using a set of predefined skills is a central challenge in robotics; solving it efficiently could enable general-purpose robots to tackle novel tasks by flexibly composing generic skills.
By Itamar Mishani, Yorai Shaoul, Maxim Likhachev
arXiv:2602.20119v2 Announce Type: replace-cross
Abstract: Solving complex long-horizon robotic tasks requires joint reasoning over abstract task structure and low-level physical interaction. While co...
By Jiahui Fu, Junyu Nan, Lingfeng Sun, Hongyu Li, Jianing Qian, Benjamin Yang, Yilun Du, Jennifer L. Barry, Kris Kitani, George Konidaris
Reward‑Guided RRT* is a sampling‑based framework that builds budget‑feasible robotic missions by iteratively selecting rewarding locations and planning motions directly in continuous state space, without a pre‑computed task‑transition graph. The method uses reward‑guided RRT* searches and a return‑feasibility check to ensure each task can be visited within the travel budget. Experiments on increasingly complex obstacle environments and task counts show solutions comparable to VNS‑PRM* while sampling far fewer states, and the framework is extended to Dubins steering and validated on a scaled car‑like robot in CoppeliaSim.
By Vin\'icius Ara\'ujo de Oliveira, Gabriel Ferreira Mendes Louren\c{c}o, Paulo L. J. Drews-Jr, Armando Alves Neto, Douglas G. Macharet