Keeping large groups of young minds engaged in the classroom can be a tall order for educational staff, because what motivates one student might not necessarily motivate the next. Working towards a future where each and every student benefits from personalized attention, MIT researchers have built an educational robot that interacts with kids and learns how to motivate them individually over time.
Dubbed Tega, MIT's robot is bright, fluffy and uses a pair of Android phones with custom software to process movement, perception and thinking. But the developers claim that what sets Tega apart from other socially assistive robots is its ability to interpret facial expressions and respond to emotions, a method known as affective computing.
This is an area we have seen other AI researchers make promising strides in of late. In 2014, Fraunhofer scientists built an app for Google Glass that detects human emotion in real time. And the tech is already making its way into some consumer-oriented products and concepts, such as Japan's popular Pepper robot and emotionally intelligent speakers.
Nevertheless, the MIT team hopes that by channeling the tech into an educational device, it can offer personalized learning companions that fine-tune their approaches to each student over time.
The researchers put Tega to work in a preschool classroom with 38 students between the ages of three and five. Each of the students worked with the robot for 15 minutes per session over a period of eight weeks, wherein they learned Spanish using a tablet computer. But rather than teaching the students, the robot assumed the role of a peer learner, offering encouragement, guidance, annoyance and boredom where appropriate.
To begin with, the robot was programmed to simply mirror the emotional responses of the students. Showing excitement when the students were excited and distraction when the students were distracted allowed Tega to monitor how these responses influenced each student's behavior. Over time, it gained an understanding of how certain cues could factor into a student's engagement and happiness and tweaked its approach accordingly to maximize learning success.
The students taking part in this personalization exercise were compared to a control group of students who received the mirroring emotions of the robot only. The researchers found that the first group demonstrated a higher rate of engagement. They also observed that certain reactions, such as a yawn or sad face, could have an immediate impact by lowering engagement or happiness.
"We know that learning from peers is an important way that children learn not only skills and knowledge, but also attitudes and approaches to learning such as curiosity and resilience to challenge," says Cythnia Breazeal, director of the Personal Robots Group at the MIT Media Laboratory. "What is so fascinating is that children appear to interact with Tega as a peer-like companion in a way that opens up new opportunities to develop next-generation learning technologies that not only address the cognitive aspects of learning, like learning vocabulary, but the social and affective aspects of learning as well."
The team says that the robot continued to fine-tune its personalized motivation approach even at the end of the eight weeks, and that a longer study would be needed to determine the ideal interaction style. They now plan to continue improving the system, which is almost entirely wireless and easily set up, by testing it out in various scenarios, including with students with learning disabilities.
Source: National Science Foundation