analiza aiqbench
This research advances hierarchical robot planning by integrating heterogeneous "black-box" skills—such as learned policies and force-controlled actions—into a unified Task and Skill Planning framework. By using Composable Interaction Primitives to bridge these diverse capabilities, the system enables robots to execute complex, long-horizon tasks while maintaining robust failure reasoning typical of traditional TAMP solvers. The approach validates this synthesis on real-world bimanual and mobile manipulators, demonstrating improved adaptability for multi-room manipulation scenarios with non-monotonic task structures.
Task and Skill Planning: Hierarchical Robot Planning with Black-Box Skills
arXiv:2504.17901v3 Announce Type: replace-cross Abstract: Task and motion planning (TAMP) is a well-established approach for solving long-horizon robot planning problems. Although TAMP methods have historically assumed that each task-level robot action, or skill, can be reduced to kinematic motion planning, recent work has explored integrating closed-loop controllers and learned skills into TAMP-style systems. Our approach integrates pre-existing, heterogeneous robot skills--including learned, force-controlled, and black-box policies--into a hierarchical planner while preserving the object-centric failure reasoning of typical TAMP solvers. We leverage Composable Interaction Primitives (CIPs) to synthesize head and tail motion plans bridging consecutive skills, facilitating both planning-time refinement and execution-time adjustment. We validate our Task and Skill Planning (TASP) approach through real-world experiments on a bimanual manipulator and a mobile manipulator, demonstrating that CIPs enable diverse robots to combine heterogeneous skills to solve complex, long-horizon tasks, including multi-room mobile manipulation problems with non-monotonic task structure.
Abstract:Task and motion planning (TAMP) is a well-established approach for solving long-horizon robot planning problems. Although TAMP methods have historically assumed that each task-level robot action, or skill, can be reduced to kinematic motion planning, recent work has explored integrating closed-loop controllers and learned skills into TAMP-style systems. Our approach integrates pre-existing, heterogeneous robot skills--including learned, force-controlled, and black-box policies--into a hierarchical planner while preserving the object-centric failure reasoning of typical TAMP solvers. We leverage Composable Interaction Primitives (CIPs) to synthesize head and tail motion plans bridging consecutive skills, facilitating both planning-time refinement and execution-time adjustment. We validate our Task and Skill Planning (TASP) approach through real-world experiments on a bimanual manipulator and a mobile manipulator, demonstrating that CIPs enable diverse robots to combine heterogeneous skills to solve complex, long-horizon tasks, including multi-room mobile manipulation problems with non-monotonic task structure.
Robotics (cs.RO); Artificial Intelligence (cs.AI)
arXiv:2504.17901 [cs.RO]
(or arXiv:2504.17901v3 [cs.RO] for this version)
Text integral sub licența deschisă a publicației · 153 cuvinte
Subiecte conexe
Hierarchical Skill Retrieval for Data-Efficient Adaptation of Vision-Language-Action Models
arXiv · 27 aug. 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · 13 oct. 2026
StageGuard: Learning Stage Transitions for Long-Horizon Robot Tasks via Agentic Distillation
arXiv · 18 sept. 2026
SurgiCLAW: An Interpretable Atomic-Skill Agent for Precision Robotic Micromanipulation
arXiv · 15 oct. 2026