EdgeMLUP-23 was an ICTP In-Person Meeting supported by the TinyML4D Academic Network and open to all. The workshop brought together educators and researchers from around the globe (47 participants from 28 countries) to develop a roadmap for sustainable university programs in embedded machine learning including the development of a common modular curriculum. This program was a collaboration led by the Abdus Salam International Centre for Theoretical Physics (ICTP), the Harvard John A. Paulson School of Engineering and Applied Sciences, Barnard College of Columbia University, Universidade Federal de Itajubá (UNIFEI), and TinyMLedu with support from Edge Impulse, Arduino, Seeed Studio, Arm, and the TinyML Foundation.
This workshop aims to develop sustainable university programs in embedded machine learning (also known as TinyML). We will do so by first bridging the gap between leveraging standard open-access course materials and adapting these materials for local and regional contexts. We will also share best practices and learnings that our network has gained through their implementation of these adaptations over the past few years. We will then augment these courses by exploring the development of university research programs centered on embedded machine learning. We will discuss how to leverage these programs to provide students and faculty with extra-curricular experience and develop new technologies and publications in the field of embedded machine learning. We will also explore how these research agendas can be used to support scientific applications and sustainable development.