Using low-cost hardware to introduce supervised machine learning in an engineering design workshop

May 10, 2024
1:00pm EST
Cummings 610
Speaker: Tanushree Burman - Quals talk
Host: Jivko Sinapov

Abstract

Quals talk:

With the rise of Artificial Intelligence (AI) systems in society, our children have routine interactions with these technologies. It has become increasingly important for them to understand how these technologies are trained, what their limitations are and how they work. Currently, much of the very limited but rapidly growing AI education introduces supervised learning topics where children learn classification concepts through a virtual setting. To encourage more hands-on learning, there have been recent efforts to incorporate AI education with robotics. However, the high cost of robotics kits is a limiting factor to many schools. As a result, there is a need for low- cost educational robotics that support AI education. We introduce a low-cost hardware tool that we call Smart Motor with both a Standalone system and a web-based user interface that we call Smart App. We also present a case study of how these interfaces support elementary school students’ understanding of ML concepts. This talk will address findings using the Smart Motor system from two pilot studies and also discuss future directions developing technology that introduces reinforcement learning concepts to students in K-12 AI education.