01
Jun
2026
Winners have been announced! 🚀
Winners have been announced. Please refer to the Winners section for more details.
Overview
The 2024 and 2025 editions built a simulation-based sim-to-real pipeline for tactile manipulation, covering tasks such as peg insertion, lock opening and visuo-tactile fusion with visuotactile sensors. However, many real-world scenarios — transparent objects, liquids, granular materials and fragile solids — remain extremely difficult to simulate and render with sufficient fidelity. Real-world demonstration data offers a powerful complement to simulation, enabling policy learning for these challenging domains.
ManiSkill-ViTac Challenge 2026 addresses this gap. Built on the ViTaMIn-B bimanual visuo-tactile data collection platform, the challenge provides real-world demonstration trajectories for contact-rich bimanual tasks. Participants train language-conditioned visuo-tactile policies directly from these demonstrations and are evaluated on the same physical hardware. By introducing language guidance, the challenge aims to advance research toward TVLA (Tactile-Vision-Language-Action) models — unifying tactile sensing, visual perception and language understanding within a single action policy.
Bridging the Reality Gap
Tasks involve transparent objects, liquids and material fragmentation — scenarios where the reality gap in simulation remains particularly wide, making real-world demonstration data essential.
ViTaMIn-B Bimanual Platform
All demonstration data is collected on ViTaMIn-B, a bimanual visuo-tactile teleoperation system that captures synchronized vision, tactile and proprioceptive streams at scale.
Language-Conditioned TVLA
Each task is paired with natural-language instructions. Policies must ground language into tactile-visual control, pushing the frontier of Tactile-Vision-Language-Action models.
Challenge Format
1
Stage 1
Registration and Screening
Feb. — Apr. 2026
A batch of sample training data for each task will be released in mid-April 2026, including wrist camera images, tactile images, wrist poses and language instructions. Participants can use this data to understand the task setup, develop initial approaches, and prepare the technical brief required for Stage 2 selection.
How to participate
- Register through the registration form — each team consists of one leader, an optional advisor, and up to four additional members
- Data format documentation and visualization tools will be provided alongside the sample data release
- A Discord server is available for Q&A and discussion with organizers and other participants
Technical Brief
- Each team submits a 2–4 page technical brief by May 1, 2026
- Content should cover team background, relevant prior work, and technical approach for the challenge
- A subset of teams will be selected by the committee to advance to Stage 2 based on the technical brief
- The number of advancing teams depends on total registrations
2
Stage 2
Real-World Robot Evaluation
May — Jun. 2026
The full demonstration dataset will be publicly released at the start of Stage 2. Advancing teams train their policies and submit them for evaluation on the physical robot platforms.
Evaluation
- Full demonstration dataset publicly released
- Each team has multiple submission attempts; each submission is evaluated on both tasks on the physical platform
- The final score is the sum of Task 1 and Task 2 performance from the same submission — the highest-scoring submission is taken as the team's result
Results and Award Ceremony
- Final rankings and results announced on June 1, 2026
- Award ceremony and winning team presentations at the CVPR 2026 Embodied AI Workshop (June 3–4, Denver)
Tasks
The challenge includes two tasks — Test Tube Cleaning and Grinding. The final score is the sum of performance across both tasks.
LeaderBoard
🥈
Baymax
2nd Place
🥇
BIT IN²Bot
1st Place
🥉
NPU-EAI
3rd Place
| Rank | Team Name | Task 1 | Task 2 | Total |
|---|---|---|---|---|
| 1 | BIT IN²Bot | 420 | 280 | 700 |
| 2 | Baymax | 200 | 40 | 240 |
| 3 | NPU-EAI | 140 | 60 | 200 |
| 4 | ZTE&BUPT | 0 | 180 | 180 |
| 5 | ResTac | 160 | 0 | 160 |
| 6 | RoboReach | 0 | 60 | 60 |
| Baseline | Official Baseline | 200 | 140 | 340 |
Baselines
Official baseline performance on the challenge tasks.
Task 1: Smash
Task 2: Tube Brush
Awards
1st Place × 1
$3,000
2nd Place × 2
$1,000
3rd Place × 3
$500
The 1st place team will be invited to present at the CVPR 2026 Embodied AI Workshop.
Challenge Timeline
The challenge progresses from website release and early data preparation to screening, leaderboard publication and award ceremony.
Jan 25, 2026
Website Release
The official challenge website becomes publicly available.
Jan 26, 2026
Workshop Acceptance
CVPR 2026 Embodied AI Workshop acceptance is announced.
Mid-April 2026
Sample Data, Documentation and Visualization Release
Sample training data, data format documentation and visualization tools released for both tasks.
May 1, 2026
Full Dataset and Baseline Release
Full demonstration dataset publicly released. Baseline leaderboard published.
May 10, 2026
Technical Brief Deadline
Deadline for teams to submit their technical brief for Stage 2 selection.
Jun 1, 2026
Final Rankings Announced
Final results and leaderboard published.
Jun 3–4, 2026
Award Ceremony
Award ceremony and winning team presentations at the CVPR 2026 Embodied AI Workshop in Denver.
Sponsors and Partners
Sudo AI
Registration
Fill in the form below to register your team. After registration, you will receive follow-up information for technical brief submission, support materials and community discussion.
Official Registration
ManiSkill-ViTac Challenge 2026 — Team Registration
Complete all fields. The team leader will receive a confirmation email after submission.
Organizers
Faculty
Rui Chen
THU
Shaowei Cui
CASIA
Jing Xu
THU
Hao Su
FDU
Student
Chaoyi Liu
THU
Rongxuan Zhang
NEU
Kaiyue Chen
XJTLU
Junlai Li
SJTU
Chenrui Zhang
NUS
Chuanyu Li
THU
Zhongyu Chen
ZJU
Lishun Zhou
Tongji University
Bailun Jiang
Wenxuan Ma
CASIA
Contact
For technical support, please join our Discord.
For other questions, please email maniskill.vitac@gmail.com.