Global Benchmark · Embodied Bimanual Manipulation

Toward a Globally Coordinated
Benchmark for Real-World
Embodied Bimanual Manipulation

Bridging the gap between lab benchmarks and real-world robot deployment

🏆 Up to $5,250 in prizes per task across three tasks — cash + purchase voucher, trophy & gift

Developer preview now live  ·  revised schedule announced soon  ·  Workshop TBD

Questions? Contact us →

Overview

Mobile manipulation, including bimanual coordination, is increasingly combined with foundation models and vision-language-action (VLA) systems. Yet the robotics community still lacks standardized benchmarks that validate these methods in real-world settings. Simulation results often fail to translate to robust deployment due to sim-to-real gaps, platform differences, and limited reproducibility. This workshop proposes a globally coordinated benchmarking initiative through a structured three-phase plan.

Phase I

Open Simulation Benchmark

Release an open simulation benchmark focusing on unstructured manipulation scenarios: cable routing, deformable material handling (thermal pad placement), and assisted living and feeding tasks. Global teams participate through an open call on a shared, reproducible leaderboard.

Phase II

Cross-Continent Real-World Validation

Teams advancing from simulation validate on real Mobile FR3 Duo robots at partner labs in Hamburg, Munich, Pittsburgh, and Shanghai. After simulation results are announced (Aug 8), qualifying teams get a hands-on bench-testing window on the real platform (Aug 10–19). Repository submissions are open now, since the developer-preview release; organizer-run simulation evaluation may begin at any time. From Aug 20–31, all real-robot bench testing and evaluation are run by the organizers — ensuring cross-site comparability.

Phase III

Workshop & Roadmap Release

Competition results, invited talks, and a panel discussion are presented at a community workshop (venue and date TBD). The workshop releases the benchmark protocol, open dataset, and community roadmap for future iterations.

Two Ways to Engage

Whether you're competing or attending, here's where to go.

The EBiM Competition

Up to $5,250 per task · cash + purchase voucher, trophy & gift
Cross-continent benchmark on Mobile FR3 Duo

Submit to the open simulation benchmark; top entries are validated on identical robots at partner labs in Hamburg, Munich, Pittsburgh, and Shanghai. Up to $5,250 in prizes per task across three tasks — cash plus a Franka Robotics purchase voucher, a trophy, and a special design gift.

  • 3 core tasks
  • 4 testbed sites
  • Standardized hardware
  • Open leaderboard
View Competition Details

EBiM Benchmark Workshop

Invited talks, panel, posters, and awards

Join us for talks from leading researchers, a panel on benchmarking standards, the competition awards ceremony, and a poster session.

  • 4 invited talks
  • Expert panel
  • Poster session
  • Awards & networking
View Workshop Program

Key Themes

Four research questions at the heart of the EBiM initiative.

Sim-to-Real Transfer

Closing the gap between simulated training and real-world robot performance.

Cross-Site Reproducibility

Ensuring results transfer across labs, hardware, and operator variation worldwide.

Foundation & VLA Model Evaluation

Rigorous benchmarking of vision-language-action models against real manipulation tasks.

Contact-Rich Manipulation

Tasks demanding precise force control, grasping under uncertainty, and deformable object handling.

EBiM Maturity Roadmap

A staged, multi-year path from first edition to long-term community standard — distinct from this year's Phase I/II/III competition timeline.

Alpha
2026 · Current
You are here · 2026

The first coordinated edition of EBiM — establishing the core tasks, simulation environments, and cross-continent real-robot validation.

Beta
2027 · Target

The next edition — broadening participation, tasks, and testbeds, and hardening protocols based on what we learn in Alpha.

Gamma
Foreseeable future · Goal

A mature, stable benchmark standard maintained as long-term community infrastructure for embodied bimanual manipulation.

EBiM follows a staged, multi-year maturity model inspired by early-access programs such as the PR2 Beta Program.

Important Dates

High-level timeline for both engagement tracks.

Competition

Benchmark Track

  • Simulation Release Preview live
  • Simulation End Aug 3, 2026
  • Results Announced Aug 8, 2026
  • Real-World Validation Aug 10 – Aug 31, 2026
  • Final Results TBD
Workshop

Workshop Program

  • Poster Submission TBD
  • Notification TBD
  • Camera-Ready TBD
  • Workshop Day TBD

For detailed timelines, see the Competition or Workshop pages.

Organizers

A multidisciplinary team spanning robotics, manipulation, and embodied AI across four continents.

Organizing Committee

Shang-Ching Liu
University of Hamburg
Donglin Chen
Northeastern University, China
Lei Zhang
Agile Robots SE
Dr. Wenkai Chen
USST, China
Yu-Hsiang Huang
National Tsing Hua University, Taiwan
Haochen Zhang
Carnegie Mellon University
Prof. Jeff Ichnowski
Carnegie Mellon University
Prof. Yi Li
Tsinghua University
Prof. Maren Bennewitz
University of Bonn
Prof. Yi-Ting Chen
NYCU, Taiwan

Advisory Board

Prof. Jianwei Zhang
University of Hamburg
Prof. Min-Chun Hu
National Tsing Hua University
Prof. Zackory Erickson
Carnegie Mellon University
Sven Parusel
Franka Robotics

Competition Support Team

Shahram Najam Syed
Carnegie Mellon University
Wen-Yu Chien
NYCU, Taiwan
Herman Prawiro
National Tsing Hua University (NTHU)
Ju Dong
University of Hamburg
Yuyang Tu
University of Hamburg
Zhentao Gao
University of Hamburg
Shiqun Qiguan
University of Hamburg
Fatma Dhaoui
Franka Robotics
Tianyi Wang
Agile Robots
Malte Huerkamp
University of Bremen
Yanxiang Zhan
University of Bremen
Jasper Güldenstein
University of Hamburg
Marian Körber
Agile Robots
Alexander Schröder
Google Cloud
Tatiana Zakharova
Franka Robotics
Xinyue Ji
Goertek Alpha Lab
Sonya Yang
AMD Research

Partners & Support

We gratefully acknowledge the partners whose support makes this benchmark initiative possible.

Gold Partners
Silver Partners

Interested in partnering? Contact us to discuss hardware support, travel grants, or community partnership opportunities.

Get Involved

Join us in shaping the future of bimanual manipulation benchmarking.

We invite teams to compete in the open simulation benchmark, and researchers to contribute poster submissions, demos, and contributed talks. All work-in-progress and position papers welcome.

Questions about submissions or participation? Contact us →