Meet Eggie: Tangible Robotics Joins the Wheeled-Humanoid Race, Betting on Dexterous Hands

The race for the wheeled, in-home robot just gained another high-profile contender. In a day of striking parallels, Tangible has emerged from stealth, revealing its robot, "Eggie," just as the industry was processing the debut of Sunday Robotics.

In a video posted to its website and X (formerly Twitter) account, Eggie—a sleek, humanoid robot with a friendly, minimalist head—is seen in a kitchen, wiping up a coffee spill from a countertop with a cloth.

The reveal, led by co-founder and CEO Bipasha Sen (an MIT PhD dropout in robotics and AI), positions Eggie as a direct competitor in the emerging category of wheeled, non-bipedal home assistants. Like Sunday Robotics, Tangible has opted for a wheeled base, prioritizing stability and efficiency over the complexity of legs.

However, a critical hardware difference is already apparent: the hands. While Sunday Robotics has made a bet on powering simple grippers, the video of Eggie clearly shows a pair of complex, anthropomorphic, five-fingered hands.

A 'Full Stack' Bet on 'Touch'

Tangible's philosophy, laid out on its website, echoes many of the "vertical integration" claims made by its new rival. The company states it is "Owning the stack" to make Eggie more capable and is building on "three pillars: dexterity, compliance, and whole-body control."

The goal is to create a machine suited for "unpredictable, cluttered spaces" and "contact-rich, real-world interactions." Where Sunday Robotics is betting its "full stack" AI can wring precision from a more simple gripper, Tangible appears to be betting that human-like hardware is essential for a world "built around humans."

This philosophy is championed by Sen, who posted in August, "What is AGI without a sense of touch? ... To me, robots aren’t just 'embodied AGI'—they are the truest form of AGI."

Data Collection via Exoskeleton

Tangible's strategy for achieving this "contact-rich" intelligence appears to rely on novel data collection methods. The company's website emphasizes that its "robots learn from real deployments" and "'In the wild' data."

A July post from Robert Scoble, who visited the company's Palo Alto offices, shed light on this. Scoble described seeing "an engineer with a partial exoskeleton on" equipped with "dozens of sensors." The purpose: "Data collection to train AI models for future robots," specifically to capture tactile information, like "how hard to press on a glass vs a carton of milk."

This aligns with Tangible's claim that "Where most robots avoid contact, Eggie thrives on it."

A Crowded Field

Tangible, backed by investors including Hubert Thieblot, joins an increasingly active field. Its focus on dexterous hands for domestic tasks places it in a different category than industrial-focused "hand" specialists like Holiday Robotics, Mimic Robotics or data-platform plays like Unitree's wheeled G1-D.

The company is candid that the home is the end goal, not the starting point. "The first step isn’t the home, it’s the technology. The AI," its website states.

To build that AI, Tangible is aggressively recruiting, advertising compensation for top research candidates between $1 million and $2 million. It's a clear signal that in the new, high-stakes race to build a true home robot, the talent war is just as intense as the hardware and software competition.