Zzapp Malaria was founded by Arnon Houri-Yafin in 2016. At the time, he was leading an R&D team at Sight Diagnostics, and as part of his job, spent three months in hospitals in India experimenting with Sight's malaria test. Witnessing the toll of the disease in person, Arnon became interested in the reason why malaria persists in many parts of the world. "It was before one of my flights to India", recalls Arnon, "that my son, who was then two years old, was running a fever. I considered canceling my trip, but my wife insisted they would be fine and that I should go. A couple of days later, I arrived at the malaria ward at the hospital in India, and realized that so many of the children admitted to the ward were the exact same age as my son. I was aware that children under the age of five were at higher risk of malaria, but this was the first time I made the connection to my own family. While I was concerned about my son possibly catching a cold, or the flu, children here were dying from a disease that was eliminated from my country long ago, back when my grandparents were my age."
Initially, Zzapp’s impact was focused on improving existing malaria control operations. Arnon led a team of software developers to create a mobile app for managing indoor residual spraying (IRS) operations. IRS involves the spraying of insecticides on the walls of houses to target the mosquitoes while resting, just after they bite, preventing transmission of malaria to the next person.
However, when reviewing the history of malaria elimination, Arnon realized that this method alone would probably not lead to elimination of the disease. Rather, another method stood out as the key in many successful elimination operations throughout the 1900’s. "Larviciding, treatment of the stagnant water bodies where mosquitoes lay their eggs, targets mosquitoes at their source,” explains Dr. Arbel Vigodny (COO) who leads Zzapp’s operations in Africa. “This method is especially important today, with the development of insecticide resistance, and with mosquitoes adjusting their behavior to bite outdoors and during the day, evading the nets and wall sprays that are currently used to fight them.”
“When looking at the countries where malaria has been eliminated, controlling the mosquito breeding sites was almost always a major component of the solution.” Houri-Yafin notes. “A notable example is the 1930’s operation led by American epidemiologist Fred Soper, who succeeded in eliminating an invading species of mosquito from Brazil within two years, by systematically applying an insecticide to every potential mosquito breeding site that could be located. Our job at Zzapp was to find a way to replicate this success, identifying and overcoming the challenges that hindered the implementation of larviciding in Sub-Saharan Africa in the past. We have more tools at our disposal than Soper did, but also more challenges. Malaria-ridden areas in sub-Saharan Africa have limited resources and a rainy tropical climate. Our goal was to find a way to operate on a low budget without compromising thoroughness.”
To address the challenges involved in the application of larviciding in sub-Saharan Africa, Houri-Yafin turned to artificial intelligence. Headed by AI director Leah Leiman, Zzapp developed a system that analyzes satellite imagery and data on climate and topography, to predict where water bodies suitable for mosquitoes may be found. “Our system tells us where to look for malaria hotspots, what measures we should implement in those locations, and how to sequence the stages of the operation”, says Leiman. “A mobile app then delivers the strategies to field workers to make sure they are implemented thoroughly”.
Alongside software development, Houri-Yafin has emphasized the importance of testing and verifying the software’s real world impact. “We're developing software using state of the art technology, but eventually it all has to work in the most difficult conditions in the field,” says Houri-Yafin. “The software has to be suitable for use in conditions of limited connectivity, and often by people without extensive backgrounds in technology. Ultimately, everything we do has to pass the field test. I like to call it 'AI with mud on its boots’.”