Palo Alto, CA, February 6th, 2026, Chainwire
ZenO opens access to egocentric audio, video, and image data captured from smart glasses and smartphones to support the next generation of physical AI systems.
ZenO today announced the launch of its public beta, opening access to a new platform designed to collect, anonymize, and structure real-world, first-person data for training physical AI systems such as robots, autonomous agents, and embodied models. ZenO will utilize Story’s Layer-1 blockchain technology as its core infrastructure.
The launch comes at a moment when physical AI is moving from research into production, but the data required to power those systems hasn’t kept pace. Robots trained on scraped web data or simulations struggle with everyday tasks that humans perform intuitively. ZenO addresses this mismatch by enabling the collection of first-person, real-world data (what people actually see, hear, and do), creating a new foundation for training embodied AI systems at scale.
By training on real-world, first-person data, physical AI models can be fine-tuned to better perceive their environments, generalize across unpredictable conditions, and perform tasks more accurately and reliably once deployed in the real world.
ZenO also recently joined a member of NVIDIA Inception, a global program designed to support startups building advanced AI technologies. Through the program, ZenO is accelerating the development of its Physical AI Data Network by leveraging access to NVIDIA’s GPU ecosystem, technical expertise, cloud infrastructure benefits, and go-to-market resources. This support enables ZenO to scale enterprise-grade, rights-cleared data infrastructure required to train robotics and embodied AI systems operating in complex physical environments.
The beta is built on ZenO’s existing MVP, now live at https://app.zen-o.xyz/, and focuses on validating the end-to-end product flow for real-world data collection: from capture to quality assurance and anonymization. The beta will run for approximately 6–8 weeks.
Unlike synthetic data or scraped online content, physical AI systems require egocentric data: what humans actually see, hear, and do in real-world environments. ZenO enables contributors to capture continuous first-person audio, video, and images using either ZenO-branded smart glasses or their mobile phones, following clearly defined data collection missions.
How the ZenO Beta Works
During the beta, users can:
- Capture real-world audio, video, and images from a first-person perspective
- Upload data through ZenO’s application for automated formatting and integrity checks
- Have data reviewed through a multi-stage QA process, including AI-based screening and human review
- Automatically anonymize sensitive information, including faces and identifiable text
After anonymization, contributors add structured metadata describing actions and environments within the footage. Approved datasets are then securely stored and cataloged within ZenO’s data marketplace infrastructure.
Incentives and Contributor Rewards
ZenO uses a two-stage incentive model for contributors:
- Immediate rewards for data collection, paid in XP during the beta phase
- Revenue sharing if the contributed data is sold, with contributors receiving a portion of downstream sales in stablecoins
This structure aligns contributor incentives with long-term data quality and commercial demand, rather than one-time labeling tasks.
Hardware and Capture Options
ZenO’s smart glasses are manufactured via an OEM partnership and launched under the ZenO brand, with specifications comparable to leading consumer smart glasses. The glasses support audio and video capture, hands-free operation, and all-day wearability. Contributors may also participate using smartphones, depending on mission requirements.
Onchain Foundations and Future Roadmap
During the beta, ZenO records wallet-signed consent and data identifiers onchain, creating a verifiable record of contributor authorization and dataset provenance. Full IP and data rights management functionality is planned for a future release.
ZenO’s long-term roadmap includes writing metadata and licensing information for user-generated datasets onto Story, enabling programmable data rights, transparent licensing, and automated revenue distribution for AI training data.
“The real world doesn’t look like the internet,” said Dawn Kim, Co-Founder of Zeno. “Physical AI systems need high-quality, rights-cleared, first-person data captured in real environments. This beta is about proving the foundation for how that data can be collected, structured, and used to train models that actually work outside the lab.”
ZenO is currently working with early data demand partners and will share traction metrics following the beta period.
To learn more or join the beta, users can visit https://zen-o.xyz.
About ZenO
ZenO is a Physical AI data collection platform focused on capturing real-world, first-person (POV) human behavior for training robotics and embodied AI systems. Through smart glasses and smartphones, ZenO enables contributors to upload video and image data generated from everyday activities. ZenO is designed to support scalable, compliant collection of real-world data for next-generation Physical AI systems.
About Story
Story is an AI-native blockchain network designed to serve as the provenance, licensing, and economic layer for AI data and models. Powered by the $IP token, Story enables datasets, models, and AI-generated outputs to be registered as intellectual property, licensed programmatically, and monetized with built-in attribution.
Backed by $136 million from a16z crypto, Polychain Capital, and Samsung Ventures, Story launched its mainnet in February 2025 and is building foundational infrastructure for the AI economy. By making IP native to the data and model lifecycle, Story provides the trust and economic rails required for AI systems to scale responsibly across enterprises, developers, and global markets.
Contact
Head Of Communications
HV
Story
henri.vies@piplabs.xyz
This article is not intended as financial advice. Educational purposes only.
