When Rebecca Ramnauth talks about ai and robots, she doesn't talk about whether they will replace human tasks or facilitate more efficient teaching and learning. Instead, she looks at the ways ai and robots can help us better connect with each other as humans.
A doctoral candidate working in the Yale Social Robotics Laboratory In Connecticut, Ramnauth has dedicated his career to building robots as tools to understand how people work. A key part of her research is studying how robots and artificial intelligence can be used to better help people in general, but specifically people with autism.
Ramnauth spoke about her work and this promising branch of artificial intelligence and robotics research as a keynote speaker at the New York City Schools tech Summit 2024 in March. In a follow-up interview, she shares what inspired his interest, the success the work has had so far, and the implications for educators.
<h2 id="using-ai-and-robots-to-help-students-with-autism-a-personal-connection-xa0″>Using ai and Robots to Help Students with Autism: A Personal Connection
Ramnauth's interest in this field was inspired by his younger sister, who was diagnosed with autism.
“I saw how he grew up and some of the difficulties he faced,” Ramnauth says. “For her, making friends is different. Things we take for granted, like making small talk, going on a date, being interrupted by an alarm, or an ambulance rushing across the street, are everyday situations, but she sees the world very differently in each of those situations. contexts.”
Inspired by her sister, Ramnauth decided to study ways in which educational tools on social assistance robots could be developed. This was an ambitious task as most research on assistive technology has focused on physical assistance. Ramnauth and her colleagues, however, are making great strides in social care.
A pilot study full of promise
One area that Ramnauth decided to investigate was eye contact, which can be something that many people with autism may have difficulty maintaining.
In a small pilot study, Ramnauth and his colleagues loaned robots to children with autism and their parents in Connecticut. “The robot did something very simple: it just looked at the child, waited until the child made eye contact, then looked at the parent, and then looked back at the child,” Reamnauth says. “The idea was that the robot would model shared attention and that, hopefully, the child would look at the robot and then turn to look at his parent and be willing to engage in some kind of conversation.”
Ramnauth and his fellow researchers found that children who received the robots engaged in more natural and spontaneous conversations with their family, and that these interactions were statistically significant.
This was notable, Ramnauth says, given that the robots were fairly basic: essentially screens with eyes that could rotate. The robots also used only basic ai technology for facial recognition.
Next steps for the investigation
This pilot study was small with only 30 participants and it is difficult to make sweeping generalizations, but the results are compelling enough to suggest future areas of research.
“It's a large enough sample to say something about how we should design the technology and that there is potential for this technology to be useful,” Ramnauth says. “These results are interesting enough that there are doctors who are starting to pay attention to this technology.”
He adds that after the success of this project, they have heard from several institutions looking to use these robots, including a pediatric center to interact with children who are experiencing a lot of stress, as well as a school in Brooklyn that works with several children with different needs. . “They are building a sensory room for their ICT classroom” he says, which will be a space where students can go and detox from the traditional learning environment. “But (they said), 'We want something that's not just toys. We want something that will engage them and light up their brain in a way that toys or books just can't.'”
Other opportunities for robots in schools
The educational implications of ai-powered robots for socially connecting with students and facilitating learning go beyond Ramnauth's specific research. Previous research has found that children respond to robots in ways similar to other social interactions, as long as the robot has eyes and physically moves in the world, Ramnauth says. The types of social interactions that robots that meet these requirements can provide are broad.
“We've done studies ranging from teaching sign language to deaf babies to teaching English to non-native English speakers, or just basic classroom instruction like teaching math or reading skills,” Ramnauth says.
Furthermore, robots can facilitate more than just learning a subject. “What I find most interesting are the social implications,” Ramnauth says. “It may encourage the quietest kid in the classroom to talk more when there's a robot in their group. Or there's a lot more teamwork. We've seen that students are much more likely to ask for help when they see a robot asking for help.”
