ZenO

ZenO | Physical AI Data Network
Building real-world data for robotics and embodied AI.

[ZenO @ZenO4AI]

https://t.co/0rI8uKm0aS

[ZenO @ZenO4AI]

Weekly Check-In is now live.

Check in once per day from Monday to Sunday and complete all 7 days to qualify for weekly XP rewards.

Stay consistent, build your streak, and earn additional rewards every week.

Check in from your Dashboard:
https://t.co/0L7kDLxOoB

[ZenO @ZenO4AI]

The full pipeline is now running, from data collection to embodied AI training.

Try it yourself.
https://t.co/uLX2zXtmCI

[ZenO @ZenO4AI]

ZenO Lab Sim Teleop is now live 🚀

You can teleoperate a simulated robot arm directly in your browser. No hardware, no setup required. Open to anyone.

https://t.co/uLX2zXtmCI

[ZenO @ZenO4AI]

All teleoperation data collected from users feeds directly into our sim robot arm training.

We are combining three types of multimodal data to enable autonomous learning: egocentric video collected through ZenO, IMU sensor data, and the new sim teleoperation data added today.

[ZenO @ZenO4AI]

And x402. Infrastructure for scalable data transactions and machine level settlement. A future where AI agents and robotics companies can access, request, and settle for ZenO data directly.
Physical AI, onchain.
Built through contribution. Growing on Base.

[ZenO @ZenO4AI]

ZenO is now LIVE on @Base.

ZenO is building a Physical AI data network for robotics and embodied AI using large-scale real-world egocentric interaction data.

Core contribution activity across the ZenO network is now being integrated onchain through Base.

Start contributing today:
https://t.co/0L7kDLxOoB

[ZenO @ZenO4AI]

With ZenO’s egocentric data pipeline and sim teleoperation, contributing to Physical AI does not require specialized equipment. People can participate from anywhere and help build the datasets that robotics and AI teams use to train real world systems.
ZenO connects industry demand for interaction data with a global contributor network.

[ZenO @ZenO4AI]

ZenO Lab is opening up sim teleoperation.

Control a robot directly in your browser to complete missions, and once you reach a certain score, your trajectory data gets saved as a training dataset. No camera, no specialized equipment needed.

ZenO opens two paths for anyone to contribute to the Physical AI industry: an egocentric video data pipeline and sim teleoperation. This mission release is the start of that second path.

Full launch within 7 days.
Join the ZenO Discord to be the first to hear.

[ZenO @ZenO4AI]

Mission #12 is now live.

Record yourself naturally opening and closing human-access doors in your daily environment.

- 3 to 15 minutes
- Human-access doors only
- No fridge, cabinet, or drawer doors
- Head or face mounted phone only
- Landscape mode required
- Both hands must stay visible
- First person POV only
- 0.5x - 0.7x wide angle

Upload via ZenO Core: https://t.co/0L7kDLxOoB
Discord: https://t.co/rBNPbIWGsx

Your data helps build the future of Physical AI.

[ZenO @ZenO4AI]

ZenO is not a data collection platform.

Anyone can collect video. That’s not the hard part.

The hard part is transforming everyday human activity into signals robots can actually learn from.

ZenO reconstructs head trajectory, reconstructs hand trajectory in 3D space, and converts egocentric human interaction into robotics-ready training data.

Teleoperation gives clean action data. But it cannot capture the messiness of real life:
uncertainty, adaptation, failure, and the intent behind every movement.
That is what everyday human activity contains.

Physical AI learns by observing humans.
ZenO turns human interaction into training data.

[ZenO @ZenO4AI]

This is exactly why we're building ZenO.

Robots like GENE-26.5 are incredible but the bottleneck has always been data. High-quality, real-world human motion data at scale.

That's what we're solving. Every person with a smartphone becomes a data contributor. Every daily action becomes training signal for the next generation of robots. 🤖

[Genesis AI @gs_ai_]

We are back. After one year of quiet building.

Introducing GENE-26.5, our first robotic brain that takes a major step toward human-level capability.

For years, robotics has struggled to learn from the world’s largest and valuable data source: Humans.

Solving it means rethinking the whole stack from the ground up:

- A robotics-native foundation model.
- A 1:1 human-like robotic hand.
- A noninvasive data collection glove for motion, force, and touch.
- A simulator that turns weeks of experiments into minutes.

GENE-26.5 is trained across language, vision, proprioception, tactile, and action. We designed a set of tasks to test how far we can go with this new paradigm.

Fully autonomous, 1x speed, one model, same weights. (Enjoy with sound on)

We are approaching the endgame for robotics.

And this is just a beginning.

[ZenO @ZenO4AI]

Turn your iPhone into a Physical AI data capture device.

Record real-world data directly through the dedicated ZenO app.

- IMU-assisted motion capture
- 6DoF trajectory reconstruction
- Depth-aware data processing
- Egocentric real-world datasets

Built for robotics and embodied AI.
Launching soon.

[ZenO @ZenO4AI]

Exactly.

Simulation scales locomotion.
Real-world data scales manipulation.

ZenO is building the physical AI data layer for that future.

[Kyle Chan @kyleichan]

Unitree founder Wang Xingxing:

In robotics, locomotion and basic motion is mostly solved. But grasping and manipulation—anything related to haptics—hasn’t been solved. That’s the key bottleneck preventing them from being deployed at scale in factories and homes.

He says that simulation is much faster for training but for manipulation tasks you still need real-world training data—for now.

[ZenO @ZenO4AI]

Mission #11 is now live.

Record yourself tidying up and organizing shoes in your daily environment.

- 5 to 20 minutes
- Head or face mounted phone only
- Landscape mode required
- Both hands must stay visible
- First person POV only
- 0.5x - 0.7x wide angle

Upload via ZenO Core: https://t.co/0L7kDLxOoB
Discord: https://t.co/rBNPbIWGsx

Your data helps build the future of Physical AI.

[ZenO @ZenO4AI]

The data Physical AI needs is changing.

For years, the focus was on locomotion and basic manipulation, walking, running, grasping. With the rise of VLA (Vision-Language-Action) models, the bar has shifted. Robots now need to see, reason about process, and predict outcomes. The same action means little without the why, the where the eyes went, and the intent behind it.

ZenO doesn't stop at collecting first-person video. When footage comes in, ZenO Studio reconstructs the head trajectory as 6-DoF pose, computes its spatial relationship to hand poses, and turns the whole thing into a format a robot can actually learn from.

Output ships in LeRobot v2.1, drop-in compatible with Physical Intelligence π0, ALOHA ACT, and similar pipelines.

Teleoperation produces clean action data, but it can't capture the environmental variability and failure modes a humanoid will actually encounter. That's why we build datasets from everyday human activity. Five minutes of someone cooking becomes a dataset where gaze, hands, and intent are baked in.

Through our app, ZenO is building infrastructure where anyone with a smartphone can produce high-quality training data. Training data for the humanoid era no longer belongs only to expensive labs.

Physical AI learns by watching us. We turn how we watch into data.

[ZenO @ZenO4AI]

ZenO Prism is live

Upload data → Earn Prisms → Unlock rewards

Even 1 upload/week is enough to start.
The more you contribute, the greater your rewards.
Built for both consistent and high-volume contributors.

Start now: https://t.co/seGP3ZKC9m

[ZenO @ZenO4AI]

Zen Signal is closing Friday, 00:00 UTC.

Thank you to everyone who posted and tagged @ZenO4AI. Your signal was heard.

All XP earned will be recorded as-is.

But this is just the beginning.
New ways to contribute are coming beyond data, beyond signal.
Stay tuned. Silence is part of the process.