A Robotic Albert Einstein Engages Students in Math and Science

IEEE member designed the digital professor to be a tutor in the home and in school

19 April 2017

Engaging students in math and science can be a difficult feat. To help, IEEE Member David Hanson developed a miniature robotic Albert Einstein that not only can play interactive games and answer questions about the subjects, but also can banter about celebrities and food. It can even gesture and stick out its tongue like the physicist did.

Founder of Hanson Robotics, a company in Hong Kong that creates lifelike robots, Hanson created Professor Einstein as his first commercial product. The digital professor is designed for students age 13 and older. Its debut at this year’s Consumer Electronics Show in Las Vegas piqued a lot of interest from consumers and the news media including TechCrunch and The Verge. The robot was the featured technology project in January on Kickstarter, raising more than the US $775,000 requested. You can preorder one for $285.

“It was crucial for us that Professor Einstein embody a sense of playfulness and wonder,” Hanson said in a news release. “He had to inspire imagination, be a great conversationalist, and make the gestures and facial expressions that the real Albert Einstein was known for.” Hanson, who holds degrees in animation and interactive art and technology, emphasizes on his website that artistry is key to bringing robots to life. He previously worked as a sculptor and technical consultant at Walt Disney Imagineering, in Glendale, Calif.

A PERSONAL TUTOR

Standing at 37 centimeters tall, Professor Einstein can walk, point, and make eye contact, and it’s programmed with 50 facial expressions. It’s equipped with Wi-Fi so that it can connect to a tablet. Users have access to the Stein-O-Matic app, which allows them to play games and work on science and math problems with the digital tutor.

Similar to virtual assistants, like the Amazon Echo and iPhone’s Siri, Professor Einstein can provide today’s weather and other helpful information. Only it puts a scientific twist on the data, asking the student to try determining the temperature sometime by counting cricket chirps, for example. The robot is programmed to share a host of fun science facts.

Professor Einstein’s accompanying mobile app is packed with experiments, brainteasers, and science-related jokes, and it is updated daily. The program’s reward system allows users to earn points, which can unlock bonus games and avatars.

To make the robot smart, Hanson relied on the same artificial intelligence software the company uses for its more advanced life-size androids. The MindCloud software processes information from millions of human-robot interactions, including from conversations and people’s responses to facial expressions, to learn and adapt in various interactions. The software is built on a deep learning model, which incorporates big data, supercomputing, and advanced algorithms to help computers make decisions. The program is open-source.

“Professor Einstein has an adaptive learning ability—which means that if you don’t understand something, he will modify the way he presents the information,” Andy Rifkin, Hanson Robotics’ CTO, says in the news release. “He has infinite patience. He will keep trying different ways of explaining things until it clicks [for the student]. Plus, because he teaches science through jokes, metaphors, and games, he makes learning genuinely fun.” Rifkin is a former senior vice president of Mattel, one of the world’s largest toy manufacturing companies, based in El Segundo, Calif., where he worked on creating interactive toys.

KEEPING IT REAL

To ensure Professor Einstein is playful and not creepy, Hanson Robotics gave it a cartoonish appearance. The company’s Sophia android, on the other hand, appears so lifelike that it has been used as an actress in a short film and was featured on the cover of Elle magazine in Brazil. But for a commercial product, Hanson wanted to avoid the “uncanny valley,” a term that refers to robots so similar to humans it becomes difficult to differentiate the two.

Learn More