X Square Robot

X Square Robot · Launch Video Breakdown: Hook, Pacing & Motion Design

Building generalist robots for real-world deployment

AI AgentsProductionSeptember 2, 2026@XSquareRobot
0:00 · The Hook · Autonomous Chemistry Experiment
0:00 / 0:00

Scene-by-scene timeline & spoken transcript

  1. The Hook

    Autonomous Chemistry Experiment

    “Twisting open a bottle cap, operating a pipette, and a scooper, coordinating both hands. This pair of robot hands is autonomously performing a chemistry experiment. Not a single piece of its training data came from a physical robot.”

    On screen
    Autonomous. 3X
    Camera
    Close-up tracking to wide static bench view
    Motion
    High-framerate stabilized robotic macro capture
  2. Problem Agitation

    The Dexterous Manipulation Dilemma

    “Behind this lies a long-standing dilemma in dexterous manipulation: collecting data through on-robot teleoperation is costly and hard to scale; collecting large volumes of data through low-cost alternatives yields data that is difficult to transfer directly to the target hardware.”

    On screen
    UMI Data Collection | Collecting data through on-robot teleoperation
    Camera
    Multi-pane split-screen montages
    Motion
    Synchronized multi-angle comparison panels
  3. Product Reveal

    Introducing TwinDEX

    “Today we introduce TwinDEX: a pair of co-designed three-finger, nine-DoF dexterous manipulation interface. One wearable for data collection, one for robot deployment. Together with an integrated data processing pipeline and policy training workflow, they form a complete collection-to-deployment system.”

    On screen
    TwinDEX | three-finger, nine-DoF dexterous manipulation interface | Dexterity | Consistency | Scalability
    Camera
    Center-locked side-by-side mirror frame
    Motion
    Typography reveal over synchronized human-robot motion tracking
  4. Feature Teaser

    Core Principle 1: Dexterity

    “The end-effector design is the result of a careful engineering trade-off. Additional fingers raise the ceiling of dexterity, but impose escalating demands on spatial packing, actuator torque density, and deployment reliability. By evaluating a broad set of typical basic human manipulation primitives, we determined that a three-finger, nine-DoF design, seven of which are actively driven, offers the best balance.”

    On screen
    Task Category | Conventional Gripper | Three-finger Gripper | TwinDEX | Human Hand
    Camera
    Macro close-ups of mechanical joints
    Motion
    Feature evaluation matrix overlay
  5. Feature Teaser

    Core Principle 2: Consistency

    “Consistency quantifies the alignment between collection and deployment devices across kinematics, contact dynamics, and visual observation. It directly determines whether robot-free data can substitute for teleoperated data in policy learning. We anchor on closed-loop deployment performance and work backward to derive quantitative requirements.”

    On screen
    Autonomous. 2X | Data Efficiency Curve | Teleoperation | TwinDEX
    Camera
    Synchronized multi-cam PIP layout
    Motion
    Animated line chart with confidence bands over synced teleop video
  6. Feature Teaser

    Core Principle 3: Scalability

    “The collection device is a wearable three-finger exoskeleton that decouples data collection from both a robot embodiment and a fixed workspace. One operator, one table, and one exoskeleton constitute a complete collection unit. The wearable form factor further provides the operator with intuitive force feedback, enabling high-quality collection of contact-rich tasks.”

    On screen
    Data Efficiency Curve | Twist a Bottle Cap | Use a Syringe | Slide Out and Flip a Notebook | Open a Toolbox | Sweep Up Trash
    Camera
    Overhead and side profile operator shots
    Motion
    Animated comparative bar chart visualization with error bars
  7. Product Reveal

    Full 24-Step Chemistry Benchmark

    “Now, back to the chemistry experiment. Twenty-four sub-actions—twisting open a bottle cap, operating a scooper and pipette, guiding liquid flow with a stirring rod, bimanual coordination across diverse fine objects, each demanding millimeter-level precision and stable force control. All data is collected from scratch through the wearable device alone, just a few hundred episodes, with zero on-robot data.”

    On screen
    Autonomous. 8X | 1/24 ... 24/24
    Camera
    Front wide shot locked on bench
    Motion
    Rapid step-counter typography synchronized to autonomous task execution
  8. Call to Action

    5.3X Efficiency & Architecture Overview

    “On a standardized benchmark covering representative manipulation primitives, TwinDEX achieves a 5.3 times improvement in collection efficiency of on-robot teleoperation while maintaining comparable data quality. TwinDEX was designed for scalable data collection from day one. As collection scales up, the system continues to support tasks of increasing complexity, making data scaling a viable path toward general-purpose dexterous manipulation.”

    On screen
    5.3 X | TwinDEX Co-designed Data Collector and End Effector | Dexterous | Consistent | Scalable | xsquare
    Camera
    Grid montage zooming out to logo reveal
    Motion
    Video array mosaic transition into typographic brand lockup

Related AI Agents Startup Launches

Explore all 1064 AI Agents launches →
OpenAI
Hook 92.0137.9M
OpenAIAI Agents

To ensure that artificial general intelligence benefits all of humanity

@OpenAI
Anthropic
Hook 92.057.6M
AnthropicAI Agents

Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use. Its capabilities exceed those of any model we’ve ever made generally available. https://t.co/2AvmEjHIX8

@claudeai
xAI
Hook 9.256.9M
xAIAI Agents

To understand the true nature of the universe.

@bot