Mosquitoes infect about 700 million people each year and cause over one million deaths, according to an estimate. A student of Vishwakarma Government Engineering College, and his guide, have developed a prototype machine that detects the presence of mosquitoes and even their type by analysing the humming noise they create by fluttering their wings. The prototype has 88.3% accuracy in detecting the type of mosquitos.

Harsh Shroff and his mentor Dr Kiran Trivedi recently won the 1st prize at the International Hackathon on Designing a Sustainable Future using ICTs, under the climate change category.


Harsh said that we started in July last year by experimenting with bird sounds and designed a device that tells you which species of bird is cooing nearby. When we succeeded, we thought of doing this on mosquitoes as vector-borne diseases like dengue, malaria and chikungunya play havoc around the world. After six months of hard work, we were able to develop the system using Machine Learning. He said work is ongoing worldwide to fight three most deadly mosquito species, Anopheles, Aedes, and Culex, 124 years after Sir Ronald Ross first made the pivotal discovery that mosquitoes cause diseases. Mosquitoes are difficult to detect manually as they are small and fly rapidly. The auditory categorisation of mosquito wing beats is used to detect them using machine learning, he said.

The duo made a proto type that collects audio data and utilises TinyML to automatically classify mosquito species. With 88.3% accuracy, the TinyML system, on the HumBug based project dataset, recognizes mosquito types. With high accuracy, the prototype has a potential to impact vector-borne illness management, he said.

Total 1,319 students participated in the hackathon from the Middle East, North Africa, Asia, Europe and USA. The average age of the participants was 27 years. The age of the oldest participant was 57 years, while the youngest was a 15-year-old. The participants submitted 32 projects, including 12 projects in the field of blockchain, 15 in climate change, and 6 in digital security.