LEGO SPIKE Prime AI Demos
As part of the SPIKE-AI team at Tufts Center for Engineering Education and Outreach (CEEO) during summer 2020, I worked closely with LEGO Education to explore possibilities to use LEGO Education SPIKE Prime Set to teach middle and high school students simple machine learning concepts. Since this is a remote internship, every demo I came up with was documented in my own room and shared via YouTube with a link.
As a background knowledge, LEGO SPIKE Prime recently launched an update which now allows users to code the educational robotics system directly with Micropython. As a result, it became relatively easy for us to write simple machine learning algorithms to be run directly on its main hub. However, the problem with Micropython is that it is not ready to handle large amount of data efficiently, which is a necessary requirement for many AI applications. Therefore, we had to work around that by either reducing the complexity of our code or using extra help from libraries that exists in a computer python environment.
Excluding the digit recognition demo which utilizes scikit-learn, all of the other demo code were written completely in the SPIKE environment, so that it could be easily replicated by anyone who has a SPIKE set, be it a middle school teacher, a high school student, or a curious LEGO enthusiasts like me :).