20924
Attempt of Training Basic-Set of Communication with Small Humanoid Robot for Autism Spectrum Disorder
One of the most important distinction on children with Autism Spectrum Disorder(ASD) is communication disability. Many of them lack eye-contact and joint attention, besides they don't pay attention and don't show interest to conversation partners. Their difficulty is responsible for a surfeit of information. They don’t know where and what to watch during conversation.
A small humanoid robot, CommU, is expected to be recognized as a friendly and less-obtrusive interlocutor of the subjects due to its cute and moderately humanlike embodiment. It can produce verbal and nonverbal signals to engage in conversation with subjects by the tele-operation method based on the predefined scripts. CommU has a possibility of resolving this problem; it has a very simple appearance, shows less emotion, deals with others always in same way, and it has an obvious novelty.
We hypothesized that after practice of conversation with robot children's basic attitude toward person would be improved.
Objectives:
The objective is to ascertain the prediction that having interaction with a small humanoid robot would improve their basic-set of communication.
Methods:
10 boys with high functioning ASD, aged between 4 and 10 (mean age=6.13, IQ≧70) were randomly assigned to two groups. Three conditions were set; Room A=Human A, Room B=Human B, and Room C= Robot(CommU). Group 1 visited three rooms in order of ‘HumanA - Robot - HumanA’, and the order of Group 2 was ‘HumanA - HumanB - HumanA’. In each room children had to talk with human or robot according to programed scenarios including communication tasks. All sessions were recorded by 3 videos which were located in different corner of the room.
Results:
First, four target behaviors were evaluated; Smile, Nodding/ Showing interest, Eye-contact, and Synchronizing with partner by listening music. Second, records of gaze following task were analyzed. About these results, Group 1 and Group 2 were compared for the purpose of excluding an effect of habituation.
There was no change in Group 2(HumanA - HumanB - HumanA) , on the other hand, in Group 1 ‘Nodding’ and ‘Synchronizing’ increased at the third session [Nodding: T(10)=-2.27, p<.05; Synchronizing: T(10)=-2.23, p<.05)]. Gaze following was increased at the second session of Group 1 [T(10)=3.07, p<.05] though Group 2 didn’t show any change.
Conclusions:
This study shows that conversation practice with robot has potential to improve the basic-set of communication of ASD. There are some factors listed; first, the novelty of robot raised their arousal level. Second, some listening skills formed because children had to adjust themselves to robot. Moreover, 70% of subjects responded that the conversation with robot was more fun and more strained. From these results, it is expected that they can face on the training with apposite feeling and arousal level.