Project Blossom wants to make social robotics accessible and approachable to all by using inexpensive materials and traditional crafting, such as knitting, crocheting and woodworking when designing the robot. The robot consists of a core mechanism with attachment points and a skin layer that can be personalized. Users can craft their own robot components and replace them easily, re-imagining what their robot may be.
The robot was designed to help children with Autism Spectrum Disorder understand subtle emotional cues in YouTube videos. A machine learning model analyzes the content and determines how the robot should behave. The robot can be used by therapists or families who wish to provide some social learning skills while the child is watching the content.
Moving forward, Project Blossom can also be adapted to many other applications, such as fan-reaction to sports videos, commentary on dramas, or even create choreographies based on online music videos. The possibilities of Blossom as as unlimited as the content on YouTube.
Project Blossom will be presented by researcher in human-robot interaction Guy Hoffman and his team at Cornell University, Greg Holman and Michael Surguitan.