Building robots that children can have long term relationships with - Part 1
One of the most common criticisms of home robots such as the Anki Vector and Petoi Bittle has been that their novelty wears away with time. As an example, in one of our previous articles, we discussed a field study in which out of 8 Anki Vector robots that were deployed in homes in Austria, less than 2 were in use after 6 months. Novelty wearing away with time has been one of the main factors holding back adoption of home robots. Many a parent has bought a robot as a Christmas gift for their child, only to find the child to get quickly bored out of it. In this context, it is very interesting to find two studies from researchers at University of Wisconsin-Madison and University of Illinois Chicago, which look into specifics of how robots can be designed so that children form long term associations with them. Both these studies will be presented at the ACM Interaction Design and Children (IDC) Conference to be held in the last week of June in Braga, Portugal.
Designing robots which can be cared for
Researchers did a study with 10 children in the age range of 8-12 to explore the idea that taking daily care of a robot such as reading to a robot before putting him to bed, tucking in the robot’s clothes and cleaning his surroundings, etc., would help build a great long term association between the child and the robot. I could relate to this study very well because I have often found my children to be very engrossed while feeding the Anki Cozmo robot with his cubes.
In the study, the children (and their families) were split into two groups. Children first read a story to the Misty II robot, which was nicknamed Micky. Half the children were asked to do activities that would form a binding with the robot (connection tasks), such as dressing up Micky, and putting a blanket on him while he was put on a charger as he went off to sleep. The other group was asked to perform chores (utility tasks) such as organizing Micky’s surroundings, putting all the toys in place, and cleaning up the room. The performance of each of the children was recorded over video. After the activity, the children provided feedback by responding to a questionnaire which was trying to measure the degree of bonding and closeness that the children formed with the robots.
After analyzing the videos, the researchers found that the children performing the connection tasks got much more involved with the robot in terms of petting the robot, wishing him “Good night”, as well as more involvement from the parents. No such bonding was seen in the group of children that performed the utility tasks.
Another interesting behavioral trend seen was that the children performing connection tasks felt much more closer and comfortable with the robot and believed that they were actually becoming friends with the robot. On the other hand, the children performing the utility tasks felt more strongly about the intelligence and sociability of the robot. One explanation of this finding is that the children who did not have to take care of the robot, perceived the robot as more independent and more like themselves. All children mentioned that they enjoyed their time with the robot, regardless of which task they were assigned.
While this is clearly an early exploratory study with a very small sample size and a very short time of evaluation, it sheds some light into design considerations that robot manufacturers might want to keep in mind while designing robots that can form long term bonds with people in the households and become more like pets instead of inane products. Clearly, designing creative ways in which children can enjoy caring for the robot is a good approach. The authors also suggest that such tasks need to be tailored to specific settings such as the age range of children, the type of activity they prefer, and their how their daily routines work. More research and product designs are definitely needed to help robots become more mainstream in our lives.