In a country where malaria remains the leading cause of death, health officials this week unveiled a new AI-powered device, its developers say, that could transform the fight against the disease, automatically tracking, identifying, and reporting on mosquito populations in real time, without a single person in the field.
The tool is a “smart” mosquito trap designed to automatically lure, capture, identify, and report on mosquito populations in real time, without the need for human monitors in the field. It is supported by the Gates Foundation in collaboration with the China Centers for Disease Control and Prevention (China CDC), and was developed by engineers from Shenzhen Galaxy Stardase Technology Co., Ltd. The entire project will run for three years. The first phase begins July 1, 2026, and will last 10 months.
According to Liya Yang, Program Engineer at Shenzhen Galaxy Stardase, the device is built to mimic a human presence to lure mosquitoes in, releasing carbon dioxide similar to human breath that attracts mosquitoes from 30 to 50 metres away, dispensing a special bait that lasts roughly a month and a half, and warming itself to mimic body heat, drawing mosquitoes close enough for an internal fan to pull them inside.
Once trapped, each mosquito is automatically photographed, and an onboard AI, described by the project team as giving the device “eyes and a brain,” identifies its species, sex, and total count without human involvement. The device also logs local weather conditions, including temperature, humidity, wind, and light, at the moment of capture.
This data, photographs, species counts, and weather readings, is transmitted automatically to a cloud-based platform, building a continuous, centralised record of mosquito activity across all deployed devices.
Beyond simple counting, the platform converts raw data into what the project team calls a “Mosquito Control Guide.” It generates a heat map showing where mosquito numbers are highest, helping field teams pinpoint where spraying or other interventions are most needed. The system also cross-references time, location, and weather data to chart mosquito behaviour patterns, giving health workers insight into when mosquitoes are most active.
The platform additionally includes an alert function: users can set custom thresholds, for example, a notification if more than ten mosquitoes are captured at a single site, or if numbers spike too quickly, and the system triggers automatic alerts accordingly. It also enables remote management, allowing operators to check whether a field device is running and switch it on or off directly from a computer.
Speaking at the event, Dr. Sulaiman Lakoh, Director of Disease Prevention and Control at Sierra Leone’s Ministry of Health, welcomed the initiative and pledged the ministry’s support in ensuring its success.
Dr. Lakoh underscored the scale of the problem the technology aims to address. Malaria, he said, remains Sierra Leone’s leading cause of death, with an even heavier toll on children under five and pregnant women, for whom it is a major cause of mortality and complications.
He explained that the Ministry’s current strategy relies heavily on vector control, particularly the distribution of bed nets through routine health services, school-based programmes, and periodic mass campaigns, but said understanding the mosquito itself is just as critical as distributing protective tools.
“Mosquito populations can shift over time, and Sierra Leone may have local strains circulating that health authorities do not yet fully understand,” Dr. Lakoh said. “It is this knowledge gap that makes the new device so relevant, by simulating the carbon dioxide and body heat that naturally draw mosquitoes to humans, the technology offers us a way to study these insects closely without direct human exposure.”
He also stressed the importance of protecting the devices once deployed in communities, noting that their value depends on remaining safe and operational long enough to deliver the surveillance data needed.
The benefits of improved mosquito surveillance may extend well beyond malaria control, Dr. Lakoh also pointed out. The same mosquito species responsible for spreading malaria are, in some cases, also vectors for other conditions present in Sierra Leone, including lymphatic filariasis, a disease that can cause the severe swelling known locally as “bigfoot,” or elephantiasis. He added that mosquito-borne illnesses such as dengue fever have also been identified in the country, though he noted that information on the latter remains limited in some respects.
The demonstration formed part of what the project team described as a core operating cycle, Auto Trapping, AI Recognition, Smart Alerts, and Precise Control, a sequence intended to move mosquito surveillance from a manual, labour-intensive process to an automated, data-driven one.
The project represents a partnership between the Gates Foundation, which is funding the initiative, and the China CDC, which is collaborating on its technical and public health dimensions, alongside Shenzhen Galaxy Stardase Technology as the device developer. For the Ministry of Health, the appeal lies in precision: rather than relying solely on broad-based interventions like bed net distribution, officials hope tools like this will let them pinpoint exactly where and when mosquito populations are surging, and respond accordingly.
As Sierra Leone continues to grapple with one of the world’s most persistent infectious disease burdens, officials say the integration of artificial intelligence and remote sensing into vector control marks a notable step toward smarter, evidence-based public health responses, one that could ultimately support not only malaria control but the surveillance of other mosquito-borne diseases as well. The project’s next phase is expected to focus on expanding the deployment of the devices to additional communities across Sierra Leone.
About the Author
