Adapting Everyday Manipulation Skills to Varied Scenarios

TL;DR

This paper presents a method for robots to adapt everyday tool-using manipulation skills to varied objects and scenarios by interpreting point clouds and adjusting motion trajectories in real-time.

Contribution

It introduces a generic encoding of manipulation skills that can be adapted dynamically using perception modules without prior object knowledge.

Findings

Successful simulation results demonstrate adaptability.

Real robot experiments validate the approach.

Robustness to object variation in manipulation tasks.

Abstract

We address the problem of executing tool-using manipulation skills in scenarios where the objects to be used may vary. We assume that point clouds of the tool and target object can be obtained, but no interpretation or further knowledge about these objects is provided. The system must interpret the point clouds and decide how to use the tool to complete a manipulation task with a target object; this means it must adjust motion trajectories appropriately to complete the task. We tackle three everyday manipulations: scraping material from a tool into a container, cutting, and scooping from a container. Our solution encodes these manipulation skills in a generic way, with parameters that can be filled in at run-time via queries to a robot perception module; the perception module abstracts the functional parts for the tool and extracts key parameters that are needed for the task. The approach is evaluated in simulation and with selected examples on a PR2 robot.