In recent years, competitions for robotic manipulation and grasping have drawn great interest from both academia and industry, such as "Amazon Picking Challenge", "Robotic Grasping and Manipulation Competition", and "Open Cloud Robot Table Organization Challenge". In addition, several benchmarks have also been proposed to help researchers better evaluate their algorithms. Some of them are based on simulators, and some are hardware benchmarks. Benchmarks use standardized testing protocol and task settings to evaluate the results of different researchers in the same way.

In this workshop, we will talk about existing benchmarks and competitions for various manipulation problems including robotic grasping, dexterous manipulation, non-prehensile manipulation, and so on. These benchmarks are composed via a close collaboration of experts in the related sub-fields. The experiences gained in the development process will also be discussed in panels for providing guidelines and inspiration for upcoming benchmarks in robotics. In addition, we will introduce a new benchmark with cloud robots, "OCRTOC benchmark" and provide a tutorial. After that, several extended abstracts will be presented to show the newest development in this field.

This workshop will be held in conjunction with ICRA 2021 and will be held online using Zoom：https://zoom.com.cn/j/86082490183

The used time zone is Pacific Time.

• 10.Apr.2021: Submission deadline for abstract

• 01.May.2021: Announcement of acceptance

• 10.May.2021: Camera ready

• 20.May.2021: Finalization of workshop program

• 04.June.2021: Workshop (online)

Invited Talks

• 08:00 - 08:05 Organizer's introduction

  • Speaker: Ping Tan, Simon Fraser University & Alibaba

  • Video: https://www.youtube.com/watch?v=KgGB_OR4VL4

• 08:05 - 08:25 Graph-Based Spatial AI

  • Speaker: Andrew Davison, Imperial College London

  • Video: https://www.youtube.com/watch?v=zzjrcaYWbo8

• 08:25 - 08:45 Learning Dexterous Manipulation on Real Robots in the Cloud - Results from the Real Robot Challenge 2020

  • Speaker: Manuel Withrich and Stefan Bauer, Max Planck Institute for Intelligent Systems

  • Video: https://www.youtube.com/watch?v=xCZyab4h504

• 08:45 - 09:05 Certified Grasping

  • Speaker: Alberto Rodriguez, Massachusetts Institute of Technology

  • Video: https://www.youtube.com/watch?v=AUsPBLoG_Qc

• 09:15 - 09:35 Potentials and Pitfalls for Benchmarking Deformable Object Manipulation

  • Speaker: Dmitry Berenson, University of Michigan

  • Video: https://www.youtube.com/watch?v=FU1ha6ECe5E

• 09:35 - 09:55 3D simulation for embodied AI: emerging research directions

  • Speaker: Manolis Savva, Simon Fraser University

  • Video: https://www.youtube.com/watch?v=Dzo52XpSnHs

• 09:55 - 10:15 Remote Experimentation of Manipulation for Online Test and Evaluation

  • Speaker: Adam Norton, University of Massachusetts Lowell

  • Video: https://www.youtube.com/watch?v=YT_1bdNV0W8

• 10:25 - 10:45 Summary of recent competitions

  • Speaker: Yu Sun (USF), Maximo A. Roa (DLR), Berk Calli (WPI), Joe Falco (NIST)

  • Video: https://www.youtube.com/watch?v=kgtpVXDXQmE

• 10:45 - 11:30 Panel discussion

  • Moderator: Hao Su, UC San Diego

  • Video: https://www.youtube.com/watch?v=J7UgZX3HeXA

• 11:50 - 12:20 OCRTOC: A Cloud-Based Competition and Benchmark for Robotic Grasping and Manipulation

  • Speaker: Ziyuan Liu (Alibaba),

  • Video: https://www.youtube.com/watch?v=9dlWUbPF2Cc

• 12:20 - 13:10 OCRTOC 2021 tutorial

  • Speaker: Minghao Gou (SJTU & Alibaba)

  • Video will be provided later

We invite interested authors to submit extended abstracts (2 pages) of relevant works. Accepted abstracts will be posted on the workshop website but will NOT be included in the official IEEE proceedings. Authors of accepted abstracts will be invited to present their works during the workshop.

All submissions should be in the form of a single PDF in ICRA format (LaTex or Word). The submission is done using Microsoft CMT portal (Submission).

Topics of interest include, but are not limited to:

• Grasping

• Motion planning

• Object pose estimation

• Manipulation

• Benchmarks or competitions

• Datasets related to robotic manipulation and grasping

For further questions please send an email to info@ocrtoc.org