Fauna Robotics unveiled Sprout, a 3.5-foot humanoid robot, on Tuesday, marking the startup’s first public demonstration of a machine designed to seem friendly rather than threatening. The robot, built over two years of secret development, stands covered in soft sage-green foam and is designed to be small enough for a 5-year-old to interact with comfortably. It represents an attempt to open a new market for robots in homes and social spaces, distinct from the industrial humanoids being developed by Tesla and Boston Dynamics.
The startup’s approach stands apart from the dominant hypothesis in robotics, which assumes humanoid robots will first be deployed in factories and warehouses before entering homes. Fauna is positioning Sprout instead as a developer platform for software engineers and researchers to experiment with, at a price of $50,000 per unit.
Design and inspiration
The design choices reflect conscious decision-making about how robots should look in shared human spaces. “Most people in this industry take inspiration from the science fiction that we grew up with,” said Fauna Robotics co-founder and CEO Rob Cochran. “I think some do so from ‘Westworld’ and ‘Terminator.’ We do from WALL-E and Baymax and Rosie Jetson.”
Sprout can walk on uneven ground, dance to classic songs like the Twist and the Floss, and grasp small objects with its grippers. The robot recovered its balance when it nearly tripped while navigating around a table during demonstrations, behavior engineers designed it to handle. It can be controlled via video game controller, smartphone application, or virtual-reality headset, and can navigate to predetermined locations within a known space.
Founders and technical foundation
The founders of Fauna—Cochran and co-founder Josh Merel, the chief technology officer—have backgrounds in AI and robotics. Merel previously worked at Google’s DeepMind, where he focused on teaching robots using AI learning techniques in simulated environments. He also co-authored a study published in Nature on an AI-powered virtual rat with fellow Fauna research scientist Diego Aldarondo.
Cochran and Merel previously worked together at CTRL-labs, a wearable neurotech company, which sold to Facebook in 2019. Cochran said he then “spent a misguided four years at Goldman Sachs” before the two decided to launch Fauna.
The hardware design drew on expertise from outside robotics. Anthony Moschella, Fauna’s vice president of hardware, previously helped design Peloton’s exercise equipment and drew inspiration from the abstract designs of Star Wars robots like R2-D2 and BB-8. He emphasized that the goal was to create a system humans would want to be around, noting that many robotics companies operate without cultural context for robot design.
Ana Pervan, a Fauna research scientist working on mapping and navigation, described the robot’s approachability. “It’s cute, and it’s not too humanoid, and I think that actually makes it a lot more fun,” she said. “It’s not verging on creepy or trying to be too human. It’s like your buddy, your pal, that’s a different thing than you.”
Market strategy and early adoption
Fauna has 50 employees and is being positioned as a “developer platform” rather than as a consumer product. Cochran believes the company is “the first American company to be actively shipping robots as a developer platform,” with the first units hand-delivered to early customers.
Early customers include Disney and Boston Dynamics. Marc Theermann, chief strategy officer at Boston Dynamics, said in a recent interview: “You take it out of the box and you can start walking it around immediately. Seeing their robot for the first time really lets you see the future a little bit. And if you squint, you can see how a robot like that would be welcomed into people’s homes.”
Challenges in the personal robotics market
The market for personal robots has historically proved difficult. iRobot, maker of Roomba vacuum cleaners, filed for bankruptcy protection last month after decades of operation. Other entrants like Anki, maker of the playful toy robot Cozmo, and Jibo, a talking speaker robot, went out of business within years of launch.
Cochran attributed the lack of success of previous attempts to the immaturity of underlying technology. “There were a lot of really brilliant attempts. I think the technology wasn’t quite there,” he said. “I do think we’re right on the precipice now where you could build a companion that is present, engaging, delightful to be around, and can also move around a space in a way that nothing ever has before.”