Embodied Future

Overview

Focus on open-source embodied intelligence and physical-world interaction, and promote open collaboration across models, data, simulation, toolchains, and real-task validation for complex dual-arm manipulation and all-terrain patrol in real industrial parks.

The Embodied Future track targets real problems in robotics and physical-world interaction. It focuses on open embodied-intelligence systems that can perceive, decide, and act, and encourages teams to build runnable, reproducible, and evaluable solutions around complex dual-arm manipulation and all-terrain patrol in real industrial parks.

The track is not about single-point demos only. It emphasizes the complete loop across models, simulation, hardware, system engineering, and task verification, helping robots complete verifiable decision-making and execution in real tasks.

Prize Information

GOAI Grand Award1 team / RMB 1M

*The winner of the Grand Award will be selected from the four track champion teams.

Champion2 teams / RMB 250K each

Runner-up2 teams / RMB 150K each

Third Prize2 teams / RMB 50K each

Additional Awards

Algorithm Innovation Award (Industry Track)1 team, RMB 30K.

Rising Team Award (Industry Track)2 teams, RMB 20K each.

Outstanding Open-source Project Award1 team, RMB 30K.

Participation AwardTeams submitting valid entries receive a competition T-shirt. Limited to 300 (ranked by work score).

Question Types

Challenge 1: General Dual-arm Collaborative Manipulation Capability Test

Based on the X-Eval simulation and real-machine integrated evaluation platform, this challenge covers VLA, WAM, dual-arm collaborative manipulation, efficient simulation, and sim-to-real transfer for algorithm development and performance evaluation.

Challenge 2: All-terrain Patrol in Industrial Parks

Based on real industrial park scenarios, this challenge focuses on autonomous navigation, real-time obstacle avoidance, multi-terrain adaptive motion control, and task scheduling to complete multi-objective inspection tasks in complex terrain.

Track Claim: Let AI Move from Digital Space into the Physical World

Enable robots to complete verifiable decision-making and execution in real tasks, and promote open collaboration across models, data, simulation, toolchains, and real-task validation.

Core Work Requirements

Entries should build runnable, reproducible, and evaluable embodied-intelligence solutions around the specified challenge, forming a complete loop from perception, decision-making, to execution and validation.
Entries should provide code, deployment instructions, third-party dependency notes, and data-source explanations, and clearly disclose the model, system architecture, runtime flow, resource usage, and open-source plan.
The preliminary round focuses on project direction, technical feasibility, and open-source value. The semi-final requires a runnable demo, code repository, technical documentation, and evaluation results. The final focuses on live demos, defense quality, implementation potential, and long-term growth.

Submission Requirements

Detailed submission requirements are as follows:

StageTimeKey Milestone / ResultKey Deliverables

Registration OpensJul 16Competition registration starts; the competition and challenges are announcedRegistration information

PreliminaryJul 16-Aug 208.20 Preliminary submission deadlinePreliminary entry, project introduction, technical proposal, demo notes, and related code/model/documentation materials

Preliminary ReviewAug 21-Aug 23Online expert review of preliminary entries; on-site finalist list confirmed; the number of advancing teams is subject to the final announcement.No additional submission unless otherwise required by the committee

Final Offline Debugging PreparationAug 25-Sep 20Finalist teams debug on real machines in groups; each group has 2-3 weeks of debugging, hosting about 3-4 teams in parallelFinal debugging materials, code/model/deployment files, demo runtime materials, technical documentation, and device-debugging notes

Grand FinalSep 22-Sep 23On-site competition and defense showcase; 5 teams compete in the morning and afternoon sessions respectively; award ceremony on Sep 23Final pitch deck, on-site demo, backup video, one-page project summary, and defense materials

*The schedule may be adjusted flexibly according to actual progress and is subject to the committee’s final notice.

Review Focus

Projects will focus on the following areas:

Challenge fit and task value
Technical solution and model capability
Engineering implementation and runnability
Evaluation results and evidence quality
Open / open-source contribution and long-term growth potential

Participant Support

This track will provide development resource support to participating teams that complete their preliminary-round submission and pass the submission validity review. The support may include computing resources, cloud services, or equivalent competition-related resources, with a maximum value of RMB 200 per team. Support will be allocated based on teams' overall preliminary-round scores, with no more than 300 allocations available per track in principle. After the preliminary-round submission deadline, the Organizing Committee will notify eligible teams via email or official competition community channels to submit the required materials, based on the submission and review results, and will distribute the resources accordingly. This resource support does not constitute a competition prize and will not affect evaluation scores or qualification for advancement. The specific forms of support, eligibility criteria, documentation requirements, distribution procedures, and timeline shall be subject to further notice from the Organizing Committee.

Best-fit Teams

This track is suitable for universities, research institutions, and enterprise teams working on embodied intelligence, robotic manipulation, VLA, world models, reinforcement learning, autonomous navigation, motion control, and system integration.

Patrol Track — Resource Access

Related open model repositories, product manuals, secondary-development API docs, and WeChat Q&A groups may be provided. Fill in the form with your GitHub account and email; materials will be sent by email.