Mukunda Bharatheesha et al. -- Plan to Win with MoveIt! - Lessons learnt from the Amazon Picking Challenge 2016

At ROSCon 2016 Mukunda gave an overview of how Team Delft took on and won the Amazon Picking Challenge in 2016. The talk provides an overview of the approach the team won, how they leveraged MoveIt, and provides incite into many lessons learned from the experience that can be used by others thinking about similar problems.

Video

Abstract

This presentation will focus on some of the key MoveIt! practices that we (motion planning team of Team Delft) followed for the Amazon Picking Challenge 2016. Particularly, the following points will be highlighted: 1) making appropriate MoveIt! API choices from a large set of options; 2) difficulties faced such as I/O synchronization with trajectories and collision checking with Octomaps and the corresponding solutions; 3) unsolved problems (mostly with robot driver) while planning around the joint limits of the robot; and 4) general recommendations for OMPL planner configurations with MoveIt!.

Slides

PDF


If you're interested in more information like this ROSCon 2017 is coming up! The call for proposals is currently open as well as registration.

Find this blog and more at planet.ros.org.


Monthly Archives

About this Entry

This page contains a single entry by Tully Foote published on May 22, 2017 2:40 PM.

Deanna Hood, William Woodall (OSRF): ROS 2 Update was the previous entry in this blog.

Matthieu Amy et al. Adaptive Fault Tolerance on ROS: A Component-Based Approach is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.