A team of researchers at Dhaka University has developed an automated, internet-connected mosquito trap that uses artificial intelligence (AI) to identify Aedes mosquitoes, the vectors of dengue, in real time.
Named MAST-CloudNet, the device attracts mosquitoes using sound, photographs captured insects and uses AI to determine whether they belong to the Aedes genus.
The trap was jointly developed by the Microsystems and Nanoengineering Lab under the Department of Electrical and Electronic Engineering (EEE) and the Mueed Research Group, also known as the Genome-Guided Vector Control and Innovation Lab, under the Department of Zoology.
The research was published on 8 October in the international journal IEEE Transactions on Instrumentation and Measurement and was funded by the Dhaka University Research Fund.
The paper, titled "MAST-CloudNet: IoT-Enabled Acoustic Trap With Real-Time AI Surveillance of Aedes", was authored by EEE Professor Syed Mahmud Ullah, Zoology Associate Professor Khandakar Ashfaqul Mueed, EEE Associate Professor Md Ahsan Habib, Zoology Lecturer Atia Sanzida Shormi, EEE students AMM Munif Khan, Fahad Arefin, Asif Rahman Dipto and Atik Bin Zakir, and Zoology students Md Siam Sarkar and Niloy Howlader.
How the trap works
Female Aedes mosquitoes produce a distinctive wingbeat sound while flying, which attracts males. The trap mimics this sound by emitting frequencies between 350 and 500 hertz, luring male mosquitoes towards the device. An internal fan then draws them in, trapping them on yellow sticky paper.
A built-in camera photographs the captured mosquitoes and transmits the images to a cloud server via the internet. An AI model analyses the images to identify the species and count the mosquitoes, with results displayed in real time on a web dashboard. This allows users to remotely monitor Aedes populations across different areas.
Although only female mosquitoes bite humans, the researchers said male Aedes mosquitoes reach adulthood earlier and remain near breeding sites. An increase in their numbers could indicate that new mosquitoes are emerging in an area, potentially serving as an early warning of dengue risk.
Tests show higher capture rates
Laboratory tests found that the sound-emitting trap captured nearly five times as many Aedes mosquitoes on average as a similar trap without sound.
The researchers then tested the device outdoors on the Dhaka University campus for three consecutive days. It remained operational despite heavy rain, strong winds and urban noise, capturing 49 Aedes mosquitoes during the trial.
Each trap costs around Tk12,000 to build, according to the researchers. By comparison, a conventional CDC light trap costs about Tk25,000, while a BG-Sentinel trap costs around Tk20,000.
A central server can process data from multiple traps simultaneously, reducing the cost of operating each unit as the network expands. The device can also run on battery power for approximately four and a half hours during a power outage.
Wider deployment planned for 2027
The researchers said the device showed promising results in controlled tests but requires further evaluation before large-scale deployment.
They plan to install 10-15 traps in Dhaka South City Corporation's residential areas in June-July 2027 to map rising Aedes populations and identify potential dengue hotspots.
The team also plans to develop a solar-powered version and integrate Long Range Wide Area Network (LoRaWAN) connectivity to transmit data over long distances without relying on Wi-Fi.
The researchers said the device has delivered promising results under controlled conditions, but further testing is needed before large-scale deployment.
Work is also underway to make the device more resistant to sun and rain.
Dhaka University Vice-Chancellor Professor ABM Obaidul Islam congratulated the team, saying the device could help identify areas with rising Aedes populations and facilitate preventive measures.
He stressed further development and effective deployment, urging researchers to improve quality of life through practical research.