IoT and Deep Learning Framework Delivers Early Dengue Detection
DU Researchers Develop AI Acoustic Trap for Aedes Surveillance
- MAST-CloudNet uses female wingbeat acoustics and cloud AI for vector surveillance
- At BDT 12,000 per unit, the device costs half of commercial optical traps
- Field deployment of 10 to 15 traps in DSCC wards scheduled for mid-2027
A multidisciplinary team of researchers at the University of Dhaka (DU) has engineered an automated, Internet-connected mosquito surveillance trap named MAST-CloudNet to monitor Aedes populations in real time. The device uses acoustic lures replicating female wingbeat frequencies to draw male vectors, captures high-resolution imagery, and transmits data to a cloud network powered by artificial intelligence algorithms for automated species classification.
Developed jointly by the Microsystems and Nanoengineering Lab (Department of EEE) and the Genome Guided Vector Control and Innovation Lab (Department of Zoology), the project was funded by the University of Dhaka Research Fund. The research was published on October 8 in the IEEE Transactions on Instrumentation and Measurement.
The multidisciplinary study was authored by Atique B. Zakir, Asif Rahman Dipto, Fahad Arefin, Niloy Hawladar, Md. Seum Sarker, A. M. M. Munif Khan, and Atia Shanjida Shormi, under the leadership of corresponding authors Prof. Dr. Saeed Mahmud Ullah, Assoc. Prof. Dr. Khandaker Ashfaqul Muid, and Assoc. Prof. Dr. Ahsan Habib.
Explaining the underlying technical architecture, the trap broadcasts an acoustic bio-lure within the 350 to 500 Hz frequency range, mimicking female Aedes wingbeats. Attracted male Aedes mosquitoes are drawn into the housing by an internal suction fan and trapped on a yellow adhesive substrate. An integrated camera module captures images of the trapped vectors and uploads them to a cloud server, where computer vision models log species counts onto a web dashboard. Because male mosquitoes emerge prior to females near breeding sites, tracking male density serves as a predictive early-warning system for dengue risk.
During laboratory evaluations, the acoustic lure captured five times more Aedes vectors compared to silent control units. In a continuous 72-hour field trial across the DU campus, the system operated reliably under adverse weather conditions, capturing 49 Aedes mosquitoes. Manufacturing costs are estimated at 12,000 BDT per unit, significantly undercutting commercial alternatives like the CDC Light Trap (25,000 BDT) and BG-Sentinel Trap (20,000 BDT). The hardware offers 4.5 hours of battery backup during grid power outages.
The researchers stated that while controlled tests yielded promising results, further evaluations are required before large-scale deployment. The team plans to deploy 10 to 15 units across residential wards in Dhaka South City Corporation during June and July 2027 to generate vector density maps. Future upgrades include integrating solar panels, LoRaWAN wireless connectivity for long-range transmission without Wi-Fi, and weatherproofing for outdoor durability.
Congratulating the team, Dhaka University Vice-Chancellor Professor ABM Obaidul Islam stated that early detection of rising Aedes populations enables swift preventive measures. He emphasized continuous development and deployment of practical research to improve public health standards.
//DBTech/IH/SMP/MIST//
