Artificial intelligence (AI) is bringing the fight to the war against dengue.
Technology cannot zap the mosquitos dead but it can be harnessed to speedily identify the species so that they can be more easily eradicated, according to Singapore-based security firm Certis.
It has developed a new AI-enabled Platform as a Service (PaaS) system that can identify mosquito species and gender – only female mosquitos bite and spread the disease – quickly by relying on a single image of an adult mosquito.
The AI system can process 1,000 images in a mere 40 seconds compared to a trained technician who will take 15 to 25 seconds to process one image.
The speed gained will quicken the time taken for remediation action like eradicating the mosquito breeding spots.
The information gained will also be combined with other data on weather, water density, vegetation and other factors to provide further insights that can help to predict future dengue hot spots.
The AI-enabled mosquito identification solution will go into action on June 1. The year-long pilot aims to further improve the accuracy of mosquito identification and thereafter to provide other information including predictive models to combat the dengue mosquitos.
As a service provider to NEA now, Certis has officers deployed islandwide in the fight against dengue. They collect mosquitos captured in Gravitraps, the special containers NEA uses to attract and trap mosquitos.
The AI-enabled mosquito identification solution is the first of its kind in the world because it integrates manpower and technology to enhance situational awareness, operational response, coordination and management, said Certis chief digital officer Fuji Foo, in a virtual media briefing May 27.
During the pilot, the PaaS, which already has information on manpower and resource assets, allows Certis to deploy field officers more quickly and effectively.
At each site, officers can use a mobile app to snap images of the insect and transmit them electronically to the lab. The AI solution then takes over with the identification process.
At the same time, by optimising how manpower is deployed, the PaaS is expected to achieve higher efficiency and expected productivity gains of up to 15 per cent, added Foo.
“At the end of the pilot, we want to enhance the productivity of our technicians in mosquito identification, analyse data better and then be able to deploy manpower more effectively.”
Certis aims to sell this solution in Hong Kong, Australia, Qatar and other countries. The company is well-positioned in this area as the fight against dengue is global.
According to the World Health Organisation (WHO), severe dengue is a leading cause of serious illness and death in some Asian and Latin American countries.
The number of dengue cases reported to WHO increased over 15-fold over the last two decades, from 505,430 cases in 2000 to over 3,312,040 in 2015. Deaths increased from 960 in 2000 to more than 4,032 in 2015.
In Singapore, NEA reported 16,000 dengue cases in 2019, adding that numbers are expected to increase this year.
CLARIFICATION at 27/05/2020 9:35pm: The article has been updated to reflect that the AI trial is conducted by Certis though not in conjunction with the NEA. Plus, it would not be deploying manpower to respond to public feedback on mosquito hotspots, as earlier reported.