Autobots

TEAM MEMBERS: Jonathan Tran , Krishan Bhalsod , Isaiah Gorman

ADVISORS: Dr. cengiz gunay , Dr. cindy robertson

TECH: mbot mega , mblock , ai , artificial intelligence

DIFFICULTY: intermediate

TARGET AUDIENCE: k-12 , college , middle-school

DURATION: 30 - 60 - 90 minutes

CURATOR(S):

PUBLISHED DATE: Apr 28 2025

CREATED IN SEMESTER: Spring 2025

Team Members:

Left to Right: Jonathan Tran, Krishan Bhalsod, Isaiah Gorman

Advisors

Dr. Cengiz Gunay
Dr. Cindy Robertson

About

Autobots introduces students to the exciting world of artificial intelligence, with a special focus on how machines learn to “see” using image recognition. Using Google’s TeachableMachine, students train their own AI models by creating unique hand gestures that control a robot’s movements. This hands-on challenge not only makes learning fun, but also shows students some of the real-world difficulties AI faces—like telling apart gestures that look very similar.

Behind the scenes, these models often rely on Convolutional Neural Networks (CNNs), a special type of AI architecture that mimics how our brains process visual information. CNNs help the computer pick out patterns, edges, and shapes in images so it can learn the difference between a peace sign and a thumbs-up. By experimenting with gesture control, students get a glimpse into how technologies like Tesla’s self-driving system work—and understand the challenges that come with training machines to interpret the visual world.