“Yes, absolutely,” says Dr. Fumihide Tanaka, a professor at Tsukuba University in Japan, when asked if he thinks telepresence robots need arms.
The current wave of commercial telepresence robots such as Beam, Ava, Vgo, Double, and Anybots are helping people who live far apart connect socially with friends and colleagues. Conspicuously, though, they do not have arms or hands, sometimes being called “Skype on wheels.” The ultra-realistic Geminoid androids, designed to be teleoperated, do have arms and hands, but they don't move.
After observing people interacting with robots for over 10 years, in multiple countries, Tanaka is convinced that robots with functional arms provide a better experience to users. “Arms and hands increase the opportunities for physical participation,” he says.
The Japanese professor has focused on challenging communication situations, such as with young children who are still learning how to navigate the social world, and interactions between individuals of different cultures.
In 2007, Tanaka visited the University of California, San Diego, to work with children between 18-24 months old. He said this age was chosen to “focus on primal forms of social interaction that are less dependent on speech.”
He found that children interacting with other children directed 53 percent of “intentional peer-to-peer contact” to the arms and hands. When he brought a robot to interact with the children, his results showed that social connectedness was correlated with the amount of touch behaviors between the child and robot.
At the Human-Robot Interaction conference earlier this year, he showed how equipping a telepresence robot with a gripper hand could help overcome social barriers.
In this study, he looked at English language learning among Japanese school children. Learning English is a common activity in Asian countries, and video conference systems such as Skype allow native English speakers to teach from abroad. However, this can be a stressful situation for students.
One major problem identified by Dr. Tanaka is “freezing.” As an English teacher explains, “When I try to say hi, they just freeze, not knowing how to respond. They tend to remain quiet and don’t talk. This type of reaction is not only seen in little kids [in Japan], but I have experienced it with young people and adults.”
To tackle this problem, Tanaka decided to put the children at the controls, giving them ability to operate the arms and grippers of the robot through a special glove.
His experiments showed that when children controlled the robot, they felt more comfortable during the session, and as a result spoke twice as many times to the English teacher, compared to the Skype-only condition. They also learned and retained words better using the robot (one example was the word "banana," which the teacher presented to the students by letting the robot grasp a real banana).
Tanaka says that the telepresence robot setup lets the children actively participate. This is the opposite situation to where the English teacher controls the robot, such as those being used in South Korea.
In a final study, Tanaka showed how a robot brought together school children in Australia and Japan. Children in New South Wales, on the east coast of Australia, took turns in controlling a humanoid robot located in Tsukuba.
“The most remarkable observation made throughout the trial was that in spite of the language difference that existed between both sides, the children were capable of communicating through the robot,” Tanaka says. “In particular, many interactions were either triggered or invoked on some physical objects that could be manipulated by the robot."
He explained that Japanese children surrounded the robot and actively tried to convey their intentions while speaking some known English words. "This type of interaction is not usually easy to induce with conventional video conference services.“
Here are some examples of the cross-continental interactions:
Together, these studies highlight something that any adult who tries to talk with children knows—playing together trumps chatting about abstract things.
It also demonstrates that if the next generation of telepresence robots are meant to connect distant family members such as grandparents and kids, something more than Skype on wheels will be needed.