'''Article:''' Interactive Robots for Communication-Care: A Case-Study in Autism Therapy by Hideki Kozima, Cocoro Nakagawa, and Yuriko Yasuda
'''Introduction to paper:'''
A small, yellow, two-ball-snowman robot with eyes (cameras) and a pointy nose, named Keepon, was used to play with typically developing and autistic children. The robot can function in autonomous or manual-control mode. The autonomous mode chooses actions based off of where the person, as well as any toy of a particular color, is located in the view of the camera. Keepon was designed to engage children in a playful way, in general - it wasn't designed to play imitation games, or anything more specific. Children demonstrated social behaviors towards Keepon after multiple sessions.
'''Application to personal research:'''
Of the three main impairments (social interaction, communication, and imagination) I've only seen experiments for the first two (no experiments designed to develop an autistic child's imagination). Maybe that would be a good area for research and experimentation.
If we created an autonomous mode for TiLAR, we could let subjects play with our robots unattended. However, since we don't want to be liable for any accidents, we can't ever really leave a subject unattended. In which case, if we're there, why not just manually control the robot?
'''Additional notes from paper:'''