Benchmarking In-Hand Manipulation

Abstract

The purpose of this benchmark is to evaluate the planning and control aspects of robotic in-hand manipulation systems. The goal is to assess the system’s ability to change the position of a handheld object using the fingers, environment, or a combination of both. We provide examples of initial and goal states (i.e., static object poses and fingertip locations) for various in-hand manipulation tasks, given an object surface mesh from the YCB dataset. We further propose metrics that measure the error in reaching the goal state from a specific initial state. When aggregated across all tasks, these metrics also serve as a measure of the system’s in-hand manipulation capability. We provide supporting software, task examples, and evaluation results associated with the benchmark.

Publication
Robotics and Automation Letters, 2020

Supplementary notes can be added here, including code and math.