Press
Reporting that features my work on TinyML and Edge AI. These pieces were written and published by others. Each entry links to the original, which is where you should read it.
IEEE Spectrum · David Berreby · 6 July 2026
A report on small language models reaching places that large ones cannot, from authenticating medicines to spotting crop disease. I was interviewed about the work we do at UNIFEI and in the AIEng4D network: classifiers for cashew plant disease, detection of malaria-carrying mosquitoes, and electrocardiogram analysis on microcontrollers. The piece describes a model I ran on a $50 Arduino board, finding mosquito breeding pools on about 3 watts.
"This is the most important area in AI nowadays. It's growing very fast."
— Marcelo Rovai, quoted in IEEE Spectrum
Science (AAAS) · Sandeep Ravindran · 20 February 2025 · Vol 387, Issue 6736
A feature on why low-power, low-cost machine learning matters in regions where electricity and connectivity cannot be assumed, reported from India, Brazil, Malaysia, Kenya and Argentina. I was interviewed for it, alongside the AIEng4D teaching network that runs TinyML courses across the Global South. The piece covers the mosquito work I did with Marco Zennaro of the ICTP: a model that tells two Aedes species apart with 98% accuracy, on a prototype whose battery lasts up to four days in the field. In Brazil, dengue affects more than a million people a year.
"I think [tinyML] is the future. It's fantastic for developing countries."
— Marcelo Rovai, quoted in Science
Read it at Science ↗ doi:10.1126/science.zdwxnqg
Headlines, quotations and links are reproduced here for reference and attribution. Copyright in each article remains with its publisher.