Dexterous Manipulation Primitives for the Real Robot Challenge

• DOI: 10.48550/arxiv.2101.11597
• Publication Date: 2021-01-01
• Authors: Claire Chen; Krishnan Srinivasan; Jeffrey Zhang; Ruoyu Zhang
• Keywords: FOS: Computer and information sciences; Computer Science - Robotics; 0209 industrial biotechnology; 02 engineering and technology; Robotics (cs.RO)

This report describes our approach for Phase 3 of the Real Robot Challenge. To solve cuboid manipulation tasks varying difficulty, we decompose each task into following primitives: moving fingers to grasp it, turning it on table minimize orientation error, and re-positioning goal position. We use model-based trajectory optimization control plan execute these primitives. These grasping, turning, primitives are sequenced with a state-machine that determines which primitive given current object state goal. Our method shows robust performance over multiple runs randomized initial positions. With this approach, team placed second in challenge, under anonymous name "sombertortoise" leaderboard. Example solving four levels can be seen video (https://www.youtube.com/watch?v=I65Kwu9PGmg&list=PLt9QxrtaftrHGXcp4Oh8-s_OnQnBnLtei&index=1).