This Lumpy Robot Lends Endearing Touch to Social Bots

Finally, a robotics design that embraces traditional crafting.

Credit: Guy Hoffman

Jokes abound about ruthless robot overlords preparing for a global takeover, but if a new type of robot named "Blossom" is any indication, the future likely also holds a place for bots that are softer and squishier than you might expect.

Blossom isn't made of sleek and shiny metal, and has no jointed appendages and no blinking lights. Quite the opposite, in fact — it closely resembles a huggable, handcrafted child's toy and looks like it could have been painstakingly stitched together a century ago.

Yes, stitched. The unusual-looking Blossom, the result of a partnership between Cornell University's Human-Robot Collaboration and Companionship Lab (HRCCL), and Google Creative Technologies Singapore, has an internal framework made of soft, flexible parts. Encasing its "skeleton" is a crocheted fabric "skin," topped with a pair of wooden "ears." [Super-Intelligent Machines: 7 Robotic Futures]

Advertisement

These handcrafted features are highly customizable, which could encourage more people to try their hand at robot-making, according to a YouTube video produced by the HRCCL that demonstrates the soft robot.

Footage of Blossom in motion shows a loosely fitted, crocheted wool sack buttoned over a shapeless, limbless body, with a slightly larger head shape perched on top. Carved wooden pieces attached to the head bob and flap in response to body movements, much as an animal's ears would. Blossom undulates in a way that seems almost organic, further distancing its design from the smooth motions and slick surfaces featured in many types of robots for home use.

"We wanted to bring back warm materials to home robotics, instead of more plastics, glass, and metal," HRCCL representatives explained in the video description.

Blossom's soft texture and fluid movements help it to appear less, well, robotic, Guy Hoffman, the robot's designer and a researcher with the Sibley School of Mechanical and Aerospace Engineering at Cornell, told the online magazine IEEE Spectrum.  

"Having worked on expressive robots for many years, one of the biggest challenges of expressive social robots is to make a rigid, hard, and digitally controlled device move in a way that seems lifelike to the viewer," Hoffman told IEEE Spectrum.

"Blossom achieves this goal in part through its physical and mechanical structure, with a lot of softness built into the materials used to drive the robot," he said.

And there could be another advantage to the imperfections or odd qualities in Blossom's looks and motion — scientists have found that people respond more positively to robots that are less than perfect, rating robots as "more likeable" when they made mistakes, according to a recent study.

One potential use for a robot like Blossom could be as a companion and teaching aid for children on the autism spectrum. Using machine learning— an application of artificial intelligence that teaches computers to learn and improve from experiences — researchers are programming Blossom to recognize and react to content in videos. Children with autism could learn to better identify certain social cues by watching videos with an "empathetic" robot, Miguel de Andrés-Clavera, Head of Creative Technology at Google Asia Pacific, told IEEE Spectrum.

"We're excited about the results that we've seen with Blossom so far, and are now looking to develop it further with partners that wish to make this social learning platform for children in the spectrum more widely available to schools and families," de Andrés-Clavera said.