An AI-based real-time surveillance tool to detect and predict mosquito-borne disease outbreaks
Problem
- Vector surveillance depends on time-consuming manual processes for larvae collection, analysis and reporting due to the shortage of entomologists.
- Widely used vector-control methods like fogging are partially effective and not species-specific. Growing insecticide resistance among mosquitoes.
- Changing weather patterns leading to vector-borne disease outbreaks in endemic and new regions.
Solution
- AI-based algorithm to identify and measure concentrations of different mosquito species based on wing beat frequencies.
- Automated AI/ML-based vector surveillance in real time for Malaria, Dengue, Chikungunya, Japanese Encephalitis and Zika.
- Data on climate and epidemiology can be combined with vector surveillance data to predict disease outbreaks and climate change hotspots, making this a Climate and Health innovation.
IMPACT
Moskeet is being used by several communities across five cities in India (Hyderabad, Thiruvananthapuram, Panjim Smart city, Vijaywada, Yanam-Puducherry), touching lives of 3,25,000 people so far.
The innovation has already shown:
- It gathers and interprets disease surveillance data 20 times faster, 3 times more accurately, and at 15% of the current cost of manual methods
- Nearly 85% savings on surveillance with value addition of real-time data;
- Reduction in fumigation costs by 20% through identification of more effective pesticides; and
- Helped focus resources on high-risk areas, thereby reducing the mosquito populations by 60% and decreasing the disease burden by 40%
- On the global stage, Moskeet was listed in the United Nations Development Program (UNDP) Solution Catalog, which is accessible by 170 country offices, where it can be considered for adoption and scaling up.
- Moskeetās pilot project has just been initiated in G South ward in Mumbai in collaboration with Municipal Corporation of Greater Mumbai . This pilot will aim at mosquito management and monitor insecticide susceptibility.
- Machine learning data sets completed for a total of 10 species of mosquitoes. The data sets have also completed field validation at Regional Medical Research Centre (ICMR affiliate), Bhubaneswar.
āMosquito-borne diseases infect 40 millionĀ people every year, and 95% of populationĀ in India resides in MalariaĀ endemic areasĀ Moskeet platform collects real-time data andĀ provides analytics for effective control of mosquito-Ā populations, disease outbreak risk analysis and pesticideĀ effectiveness. IHF support helps to expand the solutionĀ capabilities to major medically relevant mosquitoĀ species in India covering diseases like malaria, dengue,
chikungunya, filaria and Japanese Encephalitis.ā