From June 7th to 10th, 2021, three master students from the 2nd cohort of the ACEIoT master’s program namely Rosemary Nalwanga, Samson Otieno Ooko and Marvin Muyonga Ogore virtually attended the tinyML Europe, Middle East, and Africa (EMEA) 2021 Global Technical Forum (https://www.tinyml.org/event/emea-2021/) to present their preliminary Master Thesis research results on edge artificial intelligence (AI).
The forum’s focus was on Machine Learning (ML) technologies and applications capable of performing on-device sensor data analytics (technically known as inference) at extremely low power, typically in the mW range and below, and hence enabling a variety of always-on real-time use-cases using battery operated or ambient energy harvesting devices.
As great achievements, firstly the research of Marvin on offline non-invasive detection of Cholera at communal water taps has been shortlisted among the best presented abstracts; therefore getting a rare opportunity to present in the tiny talks category among the big industry industrial and research players in Edge AI technologies such as ARM, Qualcomm, Edge Impulse, Infineon, SensiML, ... Secondly, Samson was selected to present his unique research on Edge AI inference for early detection of respiratory diseases from exhaled breath during the student forum among other PHD candidates from across Europe.

Samson presenting his work during the student forum at Tiny EMEA 2021
Last but not least, the edge AI research use cases in precision agriculture, medicine and smart environment respectively explored by Rosemary, Samson and Marvin, enabled them to get selected to take part in a panel discussion about tinyML for good, interacting with top managers from ARM and Edge Impulse. This panel highlighted the need for more involvement and collaboration as a stepping stone towards ensuring tinyML for good especially in the developing worlds. It was noted that Africa features many open opportunities for TinyML applications in line with the sustainable development goals (SDG) however it is still experiencing logistic challenges in the procurement of embedded devices, components and development boards. At ACEIoT, we have temporarily relied on using embedded simulation and came up with a rapid open-source based edge AI prototyping toolstack, an approach that got a great appreciation from the forum’s audience.

Rosemary, Marvin and Samson among the TinyML for Good panelist at TinyML 20212021
The edge AI research work done by the 3 students is coordinated by Dr Jimmy NSENGA, honorary senior lecturer at ACEIoT. The students are also supervised by Dr. Didaciene Mukanyiligira, Dr. Jean Perre Muhayimpundu, Dr. Kizito, Dr. Ignace Gatare and Dr. Gerald al from the College of Science and Technology, University of Rwanda.