“If you want to be on the bleeding edge, sometimes you’re going to get cut!”

Figuratively of course! True words nonetheless from a pioneering student at Northridge High School taking on the challenge of teaching the schools GO1 Quadruped how to utilize Computer Vision. You may remember Joe Miller from a previous educator spotlight – well he’s back at it again foraying into the latest DroneBlocks course: Using Computer Vision for Object and Face Detection with the Unitree GO1. Joe and his students have been navigating the new curriculum and he has been sharing his exploits on the DroneBlocks Community Forum.

“It’s great practice for my students who are getting such a rich education in troubleshooting, problem-solving, and documentation of solutions… You can’t pay for the experience they’re getting right now.”

A pillar of the DroneBlocks ethos is to be the trailblazer in STEM curriculum and hardware development. And of course, the other pillar is to make it engaging. We believe we have achieved that with our Go1 meets Computer Vision course. By courting an MIT Robotics Researcher to explore some of the hottest topics in tech, we are delivering the cusp of technology directly into classrooms.

Here’s some further insight from the course’s creator – MIT Researcher, Gabriel Margolis.

What is the name of the degree or subject area that you are currently studying/pursuing?

I’m a PhD Candidate in the Computer Science and Artificial Intelligence Laboratory at MIT.

What was your initial inspiration for trying out Computer Vision with the Unitree GO1?

Legged robots need vision to understand their environment and interact with it intelligently. How to integrate vision with robot motor skills is an exciting open question for researchers. The Go1 offers a unique combination of capable computers, motors, and camera sensors at a low price point. This makes it an ideal platform for academic research on computer vision as well as for education.

What first interested you in a collaboration with DroneBlocks?

I think that STEM can be a powerful tool for good, and early exposure to hands-on applications is a great way to get more students excited about it. With the advent of programmable education-friendly robots in recent years, the timing is perfect for a curriculum like DroneBlocks to help educators incorporate these robots into their curriculum.

What was the most fun thing about creating this course?

Most of my time is spent writing code in service of an academic publication or creating tooling for researchers. However, I often hear from students who are just getting into legged robots and can benefit from an introductory-level resource. It was fun to create these materials with that audience explicitly in mind.

What do you hope students walk away with after taking this course?

I hope that students can have a concrete understanding of how information about the surrounding environment can be processed through the digital “eyes” and “brain” of a legged robot.