The Problem We are Solving

World's premier decentralized multi-modal data network for AGI and humanoid robotics training

OpenVision is addressing the critical challenge within the humanoid robotics industry of acquiring high-quality, diverse, and secure egocentric video training data. The market of robotics training data becomes particularly important with recent surge in demand and market interest for multimodal data

  1. Proof-of-concept and market interests for humanoid robots: recent collaboration between OpenAI and Figure, Tesla Optimus robot release, Nvidia Project GR00T general-purpose foundation model for humanoid robots, and wide use cases in healthcare and medical, companion, household, manufacturing. Many recent startups such as Physical Intelligence (spin-off from Google Robotics), k-Scale (open-source robotics team from Optimus) is an emerging category that need mass data collection.

  2. Inception of AI and AGI era: USD 200 billion market expected to grow to over USD 1.8 trillion by 2030; entering pixel-centric AI development, e.g., launch of Sora, emerging VLM

  3. Increasing market adoption of AR glasses: community hype around Apple VisionPro (~200,000 headsets sold in 10 Days), New Meta Smart Glasses; Meta’s SceneScript model to enable AI-powered AR glass

At its core, OpenVision is a two-sided platform fostering seamless creation and exchange of high-quality egocentric video training data between Data Providers with AI and Robotics companies/ institutions, through a collective data ownership governed by $VISION tokens. Three parties involved are:

  1. AI and Robotics companies/institutions who are seeking real-life, diverse, and on-demand video-based data for training their models. Ego-centric datasets labelled with actions and skill sets from the human perspective are especially valuable.

    • Examples: Cowarobot, Figure AI, Boston Dynamics, Optimus (Tesla)

  2. Data Providers who can pool in real-life data using our proprietary AR hardware. (Later on we will allow users to use their own hardware/wearables and mobile phones to capture data. )

  3. Data Validators who assess and mediate and safeguard the integrity of data submitted by reviewing and approving quality submissions.

    • Examples: AI robotics company, independent AI researchers and enthusiasts

Last updated