iRobot founder this time isn't sweeping the floor, launching AI robotic pet Familiar to keep you company and not feel lonely

Bringing 50 million Roomba vacuum cleaners into households worldwide, Colin Angle, after the iRobot acquisition by Amazon fell through and he resigned as CEO, appears at the 2026 WSJ Future of Everything Conference with his startup Familiar Machines & Magic.
(Background: The pioneer of robotic vacuum cleaners has exited the market; iRobot didn’t lose to low-cost Chinese competitors, but lives in yesterday’s ivory tower.)
(Additional context: Meta is entering the humanoid robot space! Secretly acquiring AI startup Assured Robot Intelligence, betting on “physical world interaction” leading to AGI.)

Table of Contents

Toggle

• From vacuuming to companionship, it’s more than just functions
• The graveyard of companion robots is long
• The definition of success, Angle himself explains

The 50 million Roomba vacuum cleaners have changed people’s imagination of household chores, but the problem Colin Angle wants to solve this time is much more stubborn than dust: loneliness.

On May 4, Angle officially unveiled his new company Familiar Machines & Magic at the WSJ Future of Everything Conference. Its first product, “Familiar,” is a four-legged animal robot, resembling a mix between a bear, a barn owl, and a golden retriever, about the size of a medium-sized dog.

It doesn’t speak, grasp objects, or climb stairs, but has movable eyebrows, ears, and eyes, expressing emotions through meows, purrs, and tactile fur.

From vacuuming to companionship, it’s more than just functions

In 1990, iRobot was founded under the name “Artificial Creatures Inc.” Angle spent 34 years bringing robots from industrial factories into ordinary homes. The success formula for Roomba was clear: disc shape, $200 price, dedicated task of vacuuming. Users knew what they were getting, and robots knew what to do.

Familiar’s logic is entirely different. The “companionship” feature it offers is hard to quantify, but how exactly does it keep company? Angle’s answer is: “If it gets you out of the room, going outside for a walk, that’s a real attempt to combat isolation and loneliness.”

Familiar is equipped with Nvidia Jetson Orin chips, capable of 23 independent physical movements, including head and neck turning, ears erecting, eyebrows raising, and eye tracking. All computations are done on the device, without cloud connection, and audio/video are not uploaded to servers—privacy protection is a deliberate design choice.

Co-founders Ira Renfrew and Chris Jones both come from iRobot, with backgrounds spanning Disney, MIT, Boston Dynamics, Amazon, Bose, and Sonos, covering mechanical engineering, sound design, and emotional expression simultaneously.

The long graveyard of companion robots

Familiar faces challenges beyond technology; it’s a market filled with failed precedents.

Jibo, first-generation Aibo, Vector, Amazon Astro… each once hailed by the media as a breakthrough in home robots, but ultimately either discontinued or sold poorly. Their common problem isn’t lack of technology but that users are always unsure where these devices fit into their daily lives.

An exception is Japan’s Paro, a seal-shaped companion robot mainly used in elderly care facilities, with clear use cases and target groups. Angle is also a board member of ElliQ, an elderly companion robot, so he’s well aware of the market’s difficulties.

Supporting this bet is a number: about one-sixth of the global population is affected by social isolation, which has become a public health issue; the global companion robot market is estimated to be worth $1.43 billion in 2025, projected to grow to $3.81 billion by 2032, with a CAGR of about 17.6%.

And among Americans over 68, only 9% keep pets. It’s not that they don’t want to, but as they age, caring for a real animal becomes physically and financially impractical. Familiar aims to fill this gap.

The definition of success, Angle himself explains

At the launch event, Angle made a rare statement clarifying the success criteria: “If it’s just a toy, we’ve failed. If it’s a creature you want to keep in your world, then we’ve succeeded.”

This standard is high and difficult to quantify. Roomba’s success can be measured by sales figures; Familiar’s success depends on whether it truly changes daily behaviors and reduces loneliness—metrics that can’t be verified before product launch. The earliest release is scheduled for 2027, meaning at least another year of waiting.

The next decade for household robots may not be about more powerful vacuum cleaners but about a moving, eyebrow-raising fake cat? But whether it truly solves loneliness or merely projects human desire for companionship onto another electronic device—this question still needs time to answer.