The co-founders led by their Tech lead Dr. Jimmy Nsenga have been conducting talks on the Operationalization of Research in Edge AI for Sustainable and Scalable Applied Innovation (ORESSAI) in research and AI forums. Most recently at the monthly PASET research seminar and the GIZ rwanda AI meetups.
Most of today's artificial intelligence (AI) applications using Internet of Things (IoT) data source devices rely on a cloud-centric architecture with those devices required to send collected data to the cloud for Machine Learning (ML) inference. However this architecture is not suitable for applications operating in limited communication bandwidths (like edge high-definition cameras) or subjected to strict data privacy (like patient IoT medical devices in healthcare).
An emerging technology known as Edge AI solves this data mobility challenge by enabling the Tiny ML model to data, meaning at edge, and not the other way around. As a new scientific discipline, starting to effectively research, test and prototype edge AI applications comes with some practical operationalization challenges.
In this talk, they will take you into our exploration journey of edge AI applied research, working on a diversity of edge AI applications in the domain of healthcare, environment and agriculture; using a 3-tier prototyping tool stack composed by edge Impulse, STM32IDECube and Proteus/Embedded boards. They conclude by highlighting interesting research challenges that will drive the next coming months.