Adapting Everyday Manipulation Skills to Varied Scenarios

03/07/2018
by   Pawel Gajewski, et al.
0

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.

READ FULL TEXT

page 1

page 3

page 5

page 6

research
09/13/2022

Learning Category-Level Manipulation Tasks from Point Clouds with Dynamic Graph CNNs

This paper presents a new technique for learning category-level manipula...
research
10/13/2017

Transfer of Tool Affordance and Manipulation Cues with 3D Vision Data

Future service robots working in human environments, such as kitchens, w...
research
10/14/2022

Frame Mining: a Free Lunch for Learning Robotic Manipulation from 3D Point Clouds

We study how choices of input point cloud coordinate frames impact learn...
research
09/24/2022

Learning robotic cutting from demonstration: Non-holonomic DMPs using the Udwadia-Kalaba method

Dynamic Movement Primitives (DMPs) offer great versatility for encoding,...
research
02/23/2022

Let's Handle It: Generalizable Manipulation of Articulated Objects

In this project we present a framework for building generalizable manipu...
research
11/16/2022

ToolFlowNet: Robotic Manipulation with Tools via Predicting Tool Flow from Point Clouds

Point clouds are a widely available and canonical data modality which co...
research
03/26/2021

OmniHang: Learning to Hang Arbitrary Objects using Contact Point Correspondences and Neural Collision Estimation

In this paper, we explore whether a robot can learn to hang arbitrary ob...

Please sign up or login with your details

Forgot password? Click here to reset