IEEE Computer Society Team<\/p>\n<\/p><\/div>\n
![\"TinyML](\"https:\/\/ieeecs-media.computer.org\/wp-media\/2023\/04\/05155908\/TinyML-for-the-Navajo-Nation-292x300.png\")
Last year, Vijay Janapa Reddi, associate professor in the John A. Paulson School of Engineering and Applied Sciences (SEAS) at Harvard University, was inspired to take his vision for widely accessible machine learning to students and teachers in underrepresented communities.<\/p>\n
Supported by a grant from the IEEE Computer Society\u2019s Diversity & Inclusion (D&I) Fund, Reddi partnered with colleague Brian Plancher, assistant professor of computer science at Barnard College, Columbia University, to develop a workshop for high school teachers and students in the Navajo nation. Utilizing the embedded\/tiny machine learning (TinyML) technology that they helped develop, Reddi and Plancher were able to support participants in building a small intelligent device that reacts to sounds, recognizes gestures, and distinguishes faces.<\/p>\n
\u201cThis project was designed to improve access to cutting-edge topics for students and teachers serving the Navajo Nation,\u201d reflected Reddi. \u201cWe shipped Arduino TinyML kits to our contacts at Navajo Technical University, who distributed them to the various participants.\u201d<\/p>\n
The Workshop in Action<\/h2>\n
\n
Titled \u201c2022 Edge AI Summer Institute,\u201d this workshop featured a train-the-trainer format, focusing on orienting attending teachers and students to program materials, with the goal of empowering them to educate others. Held virtually from 19 to 21 July 2022, the program drew 30 teachers and 18 students, predominantly from the states of Arizona and New Mexico, but attracting interest from around the globe. The fully remote experience allowed the team to welcome one teacher from New Jersey and one student from India who found the workshop online. Overall, the workshop brought together teachers from 26 different schools and students from 12 unique institutions, all serving the Navajo Nation.<\/p>\n
In terms of participant diversity, 50 percent of teacher attendees identified as female and 97 percent as a Person of Color (non-White), while 35 percent of students identified as female and 94 percent as a Person of Color (non-White). These demographics supported the overarching goal of bringing this education to underrepresented communities.<\/p>\n
Workshop Impacts<\/h2>\n
\n
At the conclusion of the program, teacher and student participants pointed to a positive experience. In fact, in the event\u2019s follow-up survey, nearly all teacher respondents (90 percent) indicated they planned to incorporate workshop learnings into curriculum. In addition, the overall workshop received an average rating of 4.5 out of 5 from attendees, demonstrating its impact in providing support to the target community.<\/p>\n
\u201cNow all of these schools have the ability to teach TinyML on hardware in the classroom,\u201d shared Plancher. \u201cExcitingly 90 percent of teachers who responded to our post-workshop survey said they hope to add AI\/ML\/TinyML-related content to their courses in the future.\u201d<\/p>\n
The team has posted the event recording and slides as open-access materials. In addition, these materials have been cross posted on the larger TinyML course materials page for greater visibility amongst those already leveraging TinyML resources across four continents.<\/p>\n
\u201cThis workshop was a success and should be very replicable through our open access content,\u201d indicated Reddi. \u201cWe hope that it can be used as a way to introduce students and teachers globally to the exciting world of STEM, and in particular to computer science, artificial intelligence, and machine learning.\u201d<\/p>\n
This program speaks to one of several initiatives funded by the D&I Fund in 2022. Additional results from these programs will be shared as they become available, and information on newly selected 2023 D&I Fund programs will be released shortly.<\/p>\n<\/p><\/div>\n