$\texttt{SPIN}$: distilling $\texttt{Skill-RRT}$ for long-horizon prehensile and non-prehensile manipulation

Haewon Jung, Donguk Lee, Haecheol Park, Kim Jun Hyeop, Beomjoon Kim

Keywords: Robot Skill Chaining, Imitation Learning, Planning

TL;DR: $\texttt{SPIN}$ distills a planner into a policy for long-horizon prehensile and non-prehensile manipulation by using $\texttt{Skill-RRT}$ and $\textit{connectors}$, enabling fast, robust, and zero-shot execution in simulation and the real world.

Abstract:

Current robots struggle with long-horizon manipulation tasks requiring sequences of prehensile and non-prehensile skills, contact-rich interactions, and long-term reasoning. We present $\texttt{SPIN}$ ($\textbf{S}$kill $\textbf{P}$lanning to $\textbf{IN}$ference), a framework that distills a computationally intensive planning algorithm into a policy via imitation learning.
We propose $\texttt{Skill-RRT}$, an extension of RRT that incorporates skill applicability checks and intermediate object pose sampling for solving such long-horizon problems. To chain independently trained skills, we introduce $\textit{connectors}$, goal-conditioned policies trained to minimize object disturbance during transitions. High-quality demonstrations are generated with $\texttt{Skill-RRT}$ and distilled through noise-based replay in order to reduce online computation time. The resulting policy, trained entirely in simulation, transfers zero-shot to the real world and achieves over 80\% success across three challenging long-horizon manipulation tasks and outperforms state-of-the-art hierarchical RL and planning methods.

Submission Number: 1074