Roomba Creator Colin Angle Unveils Familiar AI Pet Robot
The robotics industry just got a new player with ambitions that stretch beyond vacuuming floors. Colin Angle, the founder who put 50 million Roomba units into homes worldwide, unveiled a dog-sized companion robot called Familiar this week at the WSJ Future of Everything conference. The device comes from Angle's new company, Familiar Machines & Magic, and represents a sharp pivot from functional automation to emotional companionship.
Unlike the Roomba's straightforward value proposition—clean floors while you sleep—the Familiar is designed to form relationships with its owners. The robot resembles a cross between a bear, a barn owl, and a golden retriever, with an expressive face featuring movable eyebrows, ears, and eyes. It walks on all fours independently around the home, vocalizing but not speaking human language. Angle told The Verge that the device is a "physically embodied AI system" using generative AI via an on-device model to develop "a distinct personality."
The timing is deliberate. Angle chose May 4—Star Wars Day—to reveal a robot that aspires to hang with humans as comfortably as C-3PO or R2-D2. The choice of date signals the company's awareness that this product lives in the realm of speculative fiction becoming reality. Whether that's a selling point or a red flag depends on your tolerance for sci-fi marketing.
Technical specifications remain sparse, which is typical for early-stage hardware reveals. The Familiar will use on-device generative AI models rather than cloud-based processing, a design choice that addresses both latency concerns and privacy expectations. Angle explicitly stated he doesn't believe AI can be trusted to not say the wrong thing today, which explains the decision to avoid human speech in the first generation. The robot will follow owners from room to room, respond to gestures, and ask to be petted or taken for walks. It's an artificial life form trying to find a place in your life, according to Angle's description.
Availability timelines are sobering. The product won't be available until 2028, with consumer home testing beginning early next year. That's a two-year gap between announcement and market launch, which suggests the company is aware of the engineering challenges involved. Familiar Machines & Magic has just over 30 employees across offices in Woburn, Los Angeles, and Hong Kong. The team includes iRobot veterans plus people with experience at Amazon, Bose, Disney, and Boston Dynamics. Richard Landon, a prototype builder who created creatures for films like "Terminator" and "Jurassic Park," is also on staff.
Pricing details are equally vague. Angle said the cost will be comparable to pet ownership, which is a deliberately ambiguous benchmark. A dog can cost anywhere from a few hundred dollars for adoption to tens of thousands for purebreds, plus ongoing expenses for food, vet care, and insurance. The company isn't talking about specific numbers yet, which leaves consumers wondering whether this is a $500 novelty or a $5,000 investment.
The market for companion robots has a graveyard of failed attempts. Sony's AIBO robotic dog launched in 1999 and remains available today, but it's a niche product. Hasbro's Furby from 1998 is still sold but never achieved mainstream adoption. Jibo, a Boston startup that raised over $70 million for a countertop companion robot, shut down in 2018 when its $900 device couldn't compete with Amazon's $100 Alexa speaker. The value proposition for emotional robots has always been difficult to quantify.
Joe Jones, who helped develop the Roomba at iRobot and authored the book "Dancing with Roomba," expressed skepticism about the concept. He told MassLive that he struggles with robots designed to evoke feelings rather than perform work. "The premise of Roomba and all the other robots I developed was that the work they do determines their value," Jones said. "When the robot's 'work' is feelings, what price is reasonable?" He expects pet robots will always be hit or miss.
Steve Chambers, former CEO of Jibo, offered a more optimistic take after seeing the Familiar's introductory video. He called it "a plushie that tries to co-occupy physical space with you and acknowledge you," which he described as genuine companionship and potentially an antidote for loneliness. Chambers noted it's interesting that Angle is going from function and task with iRobot to relationship and emotion with his new venture.
The company's funding situation is somewhat opaque. Previous reports suggested Familiar Machines raised $15 million, but Angle said the total is now "decently more than that" without being specific. A 2024 SEC filing showed the company set a fundraising goal of $30 million. For context, iRobot filed for bankruptcy last year after a deal to sell to Amazon fell apart in 2024 due to antitrust concerns. The company was acquired by one of its Chinese partners, Shenzhen Picea Robotics Co.
Angle called the project "an insane thing that we've taken on, which was honestly impossible six months ago." That statement acknowledges the rapid advances in AI and robotics that made this product feasible. The launch video's narrator says Familiar Machines is "creating artificial life for a more caring world," which sounds like something out of a Philip K. Dick novel. Whether that's inspiring or unsettling depends on your relationship with technology.
What matters now is whether the engineering can match the vision. Companion robots require integrated stacks combining on-device perception, small-model natural language understanding, multimodal sensor fusion, and local or hybrid compute to meet latency and privacy expectations. The key technical questions are whether Familiar uses predominantly edge inference, what sensor suite it includes, and whether it exposes developer APIs or model-update pathways. None of these details have been published as of the May 4 announcement.
Observers will monitor whether Familiar Machines & Magic publishes technical specifications, releases SDKs or APIs for third-party integrations, and provides independent safety or privacy audits. Unit pricing and distribution partners will also affect whether a plush companion can scale beyond niche early adopters. The company has two years to prove this isn't another Jibo.
Whether consumers actually pay for emotional companionship from a robot remains the real question. The hardware costs, battery life, safety certifications, and long-term reliability are all unknowns. Angle's track record with Roomba is impressive, but cleaning floors is fundamentally different from forming relationships. Time will tell if this works—or if it becomes another expensive toy gathering dust in a closet.
Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn
Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt
Comments