Peter Wagstaff of Self Help Africa in Ireland will build an advanced machine learning algorithm that automatically analyzes high-resolution satellite images for near real-time, low-cost detection of crop pests and diseases across wide, varied landscapes. Current detection methods are either resource- or cost-intensive and limited in their ability to provide up-to-date information across large and complex geographic areas. Crop pests and diseases can alter leaf color and expose soil, which can be detected by very high-resolution satellite imaging. They will combine satellite images provided by their partner with field data on the fall armyworm crop pest collected by their project team over 18 months on smallholder plots in the Balaka district in Malawi. These data will be used to train an algorithm to detect pests and diseases. They will use cloud-based workflows to enable computationally intensive processing of large quantities of high-resolution images in near real-time. The accuracy of the algorithm will be evaluated by an independent field survey. Note: This grant is funded by the Foundation for Food and Agriculture Research (FFAR).

Cara Brook, Jean-Michel Héraud, and Soa Fy Andriamandimby of the Pasteur Institute in Madagascar, and Jessica Metcalf of Princeton University in the U.S. will establish metagenomic next generation sequencing (NGS) in Madagascar to analyze samples from undiagnosed fever patients and from bats to identify bat-derived viruses that cause human infectious diseases and help develop new diagnostics. It is estimated that up to 75% of emerging human diseases are derived from an animal reservoir. The majority of these zoonoses emerge in low-resource settings in equatorial regions likely due to living conditions and limited access to healthcare. Madagascar has long been geographically isolated, and Madagascan fruit bats are considered potential major sources of several different zoonotic diseases such as Ebola. However, identifying disease-causing viruses using traditional diagnostic methods requires the development of targeted assays and is laborious and inefficient. In contrast, metagenomic NGS is a powerful diagnostic tool that can simultaneously assay many viruses and identify new ones. Their institution serves as a national reference laboratory for several diseases in Madagascar, and to help develop local next generation sequencing and data analysis capacity, they will pilot the approach with a subset of 380 human clinical samples already collected. They will then process an additional 410 human samples for NGS library preparation, sequencing, and pathogen analysis in Madagascar.

Carol Sze Ki Lin of the City University of Hong Kong and Srinivas Mettu of the University of Melbourne in Australia will develop a new, low-cost bioreactor system to mimic the human gut and facilitate simultaneous growth of multiple bacterial strains with diverse growth requirements. A healthy mixture of bacteria in the human gut is essential to overall health, and live biotherapeutics could be used to restore this population in infants whose gut microbiota has been damaged by malnutrition. Manufacture of these therapeutics is difficult and expensive: the human gut contains multiple strains of bacteria with diverse environmental and nutritional growth requirements, and designing a bioreactor to accommodate such variation is difficult. They will create stratified growth zones within one reactor based on immobilization of the bacteria on a low-cost, biodegradable plant-based cellulose hydrogel. They will alter the porosity and surface chemistry of the hydrogel to create variations in pH and oxygen and nutrient levels within the reactor to simulate the human gut from stomach to rectum. This will facilitate the simultaneous growth of ten bacterial strains with diverse growth requirements, first in a lab setting and later on a commercial scale.

Nguyen Thanh Hung and colleagues in Children’s Hospital 1 in Vietnam will implement next generation sequencing to identify the diverse viral causes of encephalitis in children in Vietnam and develop more accurate and rapid diagnostics to improve clinical outcomes. Encephalitis is an inflammation of the brain commonly caused by viral infection and is a major contributor to childhood morbidity and mortality worldwide. Treatment requires rapid diagnosis so that the appropriate antimicrobial therapy can be administered. However, traditional diagnostics are inadequate, and identify less than half of cases. Further confounding accurate diagnosis is the emergence of new and diverse pathogens that cause encephalitis. Next generation sequencing could overcome these challenges by more rapidly sequencing viral nucleic acids than previous methods and at lower costs. They will test this at their 1,600-bed hospital located in Ho Chi Minh City in Vietnam with samples from around 150 children presenting with encephalitis and compare the results with routine diagnostics. Overall, they aim to identify the major causes of encephalitis across Vietnam, map them temporally and spatially, and characterize disease outcomes.

William Kunin of the University of Leeds in the United Kingdom will develop methods to monitor agricultural pest outbreaks in Africa using data from dual-polarization weather radar. Pest infestation is responsible for up to 50% of pre-harvest crop loss in Central Africa, and control depends on the ability to monitor local pest outbreaks and movement over large areas – a difficult and expensive task. Sophisticated dual polarization Doppler weather radar is designed to detect airborne objects like rain and hail. However, because it is sensitive to size and shape, it can also be used to detect specific types of insects, including crop pests, and algorithms exist to separate the different signals. They will use micro-CT scans to create three-dimensional models of pests common to Rwandan crops and conduct simulations to predict radar data patterns for swarms of particular pests at varying densities. Through collaboration with the Rwandan meteorological service, they will analyze radar data to identify pest outbreaks and their movements across the country. Once established, the methods will be applied across Africa to provide a low-cost, advanced warning system for crop protection.

Nico Vandaele and Catherine Decouttere from the Katholieke Universiteit Leuven in Belgium along with workers at the St. Francis of Assissi Community Dispensary in Kenya will develop a decision support tool for community workers who vaccinate children that incorporates the diverse characteristics and needs of the local parents and caregivers to help them improve vaccination uptake. The diverse characteristics of caregivers in a specific catchment area, such as their profession and household situation, affect the challenges they face when seeking health care and the solutions needed to overcome them. They plan to develop a tool for community workers to map these characteristics so that they can exploit them to provide a better health service, such as by tailoring appointment times, arranging transport, and sending reminders. This will enable them to work more efficiently. They will build a paper-based prototype in collaboration with community workers and pilot test the tool and validate maps of selected communities.

Muhammad Imran Nisar, Furqan Kabir, Fyezah Jehan, and Syed Asad Ali of Aga Khan University in Pakistan will employ a metagenomic sequencing approach to better identify bacterial and viral infections, including those caused by novel pathogens, in newborns and infants in Pakistan. Pakistan has the highest neonatal mortality rate in the world. Almost one third of neonatal deaths are thought to be caused by an infectious disease, but accurate diagnosis is challenging in low-resource settings. In a recent study, despite using elaborate detection methods, only 28% of suspected infectious cases could be confirmed. In addition, novel pathogens escape detection because conventional microbiological methods are limited to detecting a predetermined selection. To overcome these limitations, they will establish a metagenomics approach for next generation sequencing with the cloud computing IDseq software to identify the infectious pathogens. They will sequence and analyze 150 archived blood and nasopharyngeal specimens from a previous study and additionally analyze blood and swab samples to be collected from sick newborns and infants with suspected infections over a six-month period at a primary healthcare center in a peri-urban community in Karachi, Pakistan.

Ricardo Valladares of Siolta Therapeutics in the U.S. will develop a low-cost method to manufacture large quantities of mixed populations of bacteria for use as biotherapeutics to restore a healthy population of gut microbes in infants. A diverse population of bacteria in the infant gut is essential for health, but malnourishment and antibiotics can destroy microbial diversity and cause metabolic and immune problems. Gut health may be restored by treatment with a consortium of bacterial strains. However, mass production of these strains is challenging as many have complex nutritional and environmental growth requirements. To date consortium microbes have been cultured individually, resulting in a costly, small-scale process. They will first study bacterial populations in healthy infant guts to determine what species are required and the combinations that co-exist naturally. To support their growth in vitro, they will optimize a vegetable-based culture media composed of low-cost, widely available food staples, and then apply bioreactor technology used by commercial breweries to manufacture large quantities of the microbial mixture at low cost. Together, these approaches could help make microbiota-based biotherapeutics accessible in low- and middle-income countries.

Benedict Mongula of the University of Dar es Salaam in Tanzania will analyze two apparently conflicting national agricultural policies centered on either large-scale agriculture or smallholder farmers and determine how to combine them to benefit all stakeholders for inclusive agricultural transformation. Agriculture is central to the Tanzanian economy, yet its impact is limited by a lack of infrastructure, education, and market access. Current agricultural policy is shaped by two conflicting approaches: large-scale agriculture under the Southern Agricultural Growth Corridor (SAGCOT), driven by wealthy foreign investors; and smallholder farming under the government-led Agricultural Sector Development Programme (ASDP). SAGCOT has been criticized for removing land from rural farmers and increasing poverty, while ASDP has many projects in place but has not demonstrated significant impact on the industry. They will investigate how to link the two approaches for inclusive agricultural transformation – improvements to the industry that protect all stakeholders, regardless of social class, culture, gender or age. Through consultation with participants, including government officials, investors and ordinary citizens, they will analyze the two approaches, identify which of the large-scale agricultural investments are most consistent with inclusive transformation, and determine to what extent both SAGCOT and ASDP address the needs of the diverse population. The new policy framework will be presented in the form of a journal article, stakeholder meetings, and workshops.

Melanie Bannister-Tyrrell of Ausvet in Australia will create an SMS-based communication system for farmers in Kenya to anonymously report crop disease and pest infestations and generate surveillance data to minimize crop loss. Pest infestation and disease cause substantial crop losses each year. In many low-income countries, farmers do not report disease to local agricultural authorities because they fear their crops will be destroyed without compensation. Yet information on the presence and spread of pests is needed to inform decisions on planting and control measures. They will optimize the design of an SMS interface for anonymous reporting using SMS codes for diseases and pests. The system will be piloted in one county, leveraging existing users and community volunteers to recruit new users. Reports will be used to estimate the local prevalence of pests and disease using modified small-area statistics. If a credible threat is detected, all users will be alerted. The system will improve connection and trust between farmers and agricultural extension workers and improve surveillance.

Mustafa Naseem of the University of Michigan in the U.S. will create an android application to present digital immunization and performance data from front-line health workers to their medical supervisors to improve vaccination coverage in Pakistan. Polio is a vaccine-preventable disease, eradicated in much of the world yet endemic in Pakistan due to poor compliance with immunization schedules. Vaccine administration in rural provinces is challenging because of understaffed, understocked, and sparsely-located healthcare centers. To improve this, the government introduced eVaccs – a smartphone-based monitoring system to track the movements and vaccinations administered by each vaccinator. However, these data remain largely inaccessible to their direct supervisors. To address this, they will develop an application to present them with relevant data in a useful format in real time. This will enable supervisors to better monitor performance, identify key challenges to comprehensive vaccination coverage, and help them better manage vaccine supplies. They will perform behavioral experiments in the field to test whether their approach positively influences vaccination coverage.

David Mills of the University of California, Davis in the U.S. will determine whether specific plant-based oligosaccharide formulations can drive mixed-culture growth of selected strains of intestinal bacteria for the low-cost and efficient production of live biotherapeutics. Microbial colonization in the human gut is important for overall health. It has been shown that oligosaccharides can provide a food niche to specifically enrich key colonizing bacteria, even in the competitive environment of the human gut. They will exploit this to grow multiple strains simultaneously in a controllable, scalable manner. They have recently developed analytical tools to characterize over 1,000 plant polysaccharides. These will be screened using bioinformatics methods and then in vitro to identify optimal oligosaccharide-therapeutic bacteria combinations that can support mixed-culture growth. They will then progress to bioreactor screening of the top candidate combinations. Once established, the live biotherapeutics will be formulated with their paired oligosaccharides for synbiotic application that may enable them to more readily colonize the human gut.

Caroline Kabaria of the African Population & Health Research Center in Kenya will use geographic information systems (GIS) to map the location of health facilities and community health volunteers in Kenya to identify particularly marginalized slum populations that need better access to health services such as vaccinations. Nairobi and Kisumu contain over 100 slums where residents live in dense and unsanitary conditions. The specific health needs of these residents are difficult to assess from national statistics that often exclude them. To address this, they will conduct an in-depth assessment to identify equity gaps specifically in childhood immunizations. Community health volunteers will be trained to use GIS mapping techniques and to register households, and the data will be integrated with the existing district health information system (DHIS 2). This will be used to produce an interactive map of the two cities that includes the spatial and social structures of informal settlements and the location of health facilities. They will also provide training and guidance to local stakeholders on how to utilize the map to improve vaccination coverage.

Aart Van den Beukel of Safi Sana in the Netherlands will enable digital monitoring of the entire waste and sanitation supply chain to improve quality control, reduce costs, and help communities transform waste into resources such as agricultural and energy products. Poor sanitation and waste management can drive poverty and disease, but it is difficult to monitor in low-resource settings. The sanitation and waste industry can also provide unique opportunities for social and economic benefit when communities are given access and support. They will install data collection systems to monitor all elements of waste processing such as GPS data from waste truck drivers to help improve efficiency and reduce contamination and costs. These data will be provided on a tailored, computer-based platform to relevant stakeholders and local communities, along with education and awareness training to encourage new waste-converting enterprises.

Hidenori Harada of Kyoto University in Japan will develop and test a largely automated system for the regular removal and safe disposal of fecal sludge from septic tanks in Asia. Billions of people rely on septic systems to collect human waste. Regular emptying and proper disposal of fecal sludge are critical to avoid potential contamination of clean water by backups, leaks and illegal dumping of waste. Hazardous sanitation problems occur when customers are unaware of the need to have their system emptied or when truck operators hired to empty sludge illegally dump the waste outside of a proper treatment site. They will institute a program by which tanks are – at no charge to the customer - equipped with tagged covers that provide access for safe emptying and enables them to be registered with a centrally-monitored system that will automatically notify customers when their tank needs to be emptied. Payment by households for emptying the tanks will be collected electronically, and the sludge will be removed and transported via trucks equipped with GPS to ensure proper disposal of the waste at a treatment center. The system will be tested in cooperation with a public emptier in Mandalay, Myanmar.

Paul Namwanja of the Community Health Centre, Busabala in Uganda will implement mobile technology for village health teams to monitor households and childhood vaccinations on remote islands in Uganda and to establish public-private partnerships with community health workers and boat owners to improve vaccine coverage. Vaccination rates on Ugandan islands are significantly lower than the national level because of unreliable transportation to access mainland healthcare centers; island residents rely on commercial fishing boats to travel and face long wait times for healthcare service. They will modify an existing mobile application for village health teams to create a registry of new and expectant mothers by recording household location and required vaccines. These data can then be shared with mainland health workers and used to facilitate access to a subsidized boat taxi service for travel to medical appointments. They will test their approach on six trial islands by providing the mobile phones and application along with training to health workers and recruiting boat drivers and evaluate its effect on vaccination coverage.

Rumi Chunara of New York University in the U.S. will collect data from mobile phones of healthcare workers to develop algorithms that will help prioritize healthcare resources to increase vaccine coverage in Punjab, Pakistan. Immunization is one of the most cost-effective and successful public health strategies and is estimated to prevent up to three million deaths each year. Still, many rural areas of low- and middle-income countries have an under-vaccinated population due to a lack of formal education and awareness of the importance of vaccinations. They will collect data on disease incidence and vaccine resources and coverage from the mobile phones of 3,800 rural healthcare workers. These data will be used to train an artificial intelligence model to identify the ideal times and locations to target vaccine efforts. The model will also address accessibility and awareness issues by incorporating distance to healthcare centers and frequency of visits. Their approach will be evaluated by analyzing coverage before and after implementation. Once proven, it can be scaled to other areas in Pakistan.

Muhamed Bizimana of CARE in the U.S. will establish an electronic system to group urban slum households and connect them with sanitation services and financing options to reduce unsafe dumping of human waste in urban Cote D’Ivoire cities. Sixty percent of urban households in Cote D’Ivoire have unsafe sanitation systems – septic tanks improperly emptied, public dumping of human waste, and open defecation – which causes a serious health threat and limits economic development. A goal has been set for all households to have access to clean drinking water and sanitation by 2030. To achieve this requires behavior change. They will set up an electronic community sanitation support system in the trial cities of Bouake and Katiola that facilitates pooling of households’ resources to acquire better toilets and more affordable waste disposal. The system will group households geographically, establish a financial scheme for sanitation equipment and services, and link groups with waste removal service providers. They hope to cause a domino effect on behavior change towards safe sanitation habits, with plans to expand across the country by collaboration with the national office of sanitation and local municipalities.

Rajiv Rimal of George Washington University in the U.S. will increase vaccination uptake in Nepal by improving the state of health centers to make them more enjoyable and practical places for mothers to bring their children and to motivate health workers to provide better care. In Nepal and many other low-resource settings, essential health services such as vaccinations are often provided in settings with long waiting times and limited facilities, which is also demotivating for the staff. To address this, they will interview recent mothers and service providers to understand the challenges around receiving and providing vaccinations. These will be used to design an intervention that involves improving the relationship between service providers and those giving the vaccinations as well as cleaning up the health centers by painting and providing seating. They will test the ability of their approach to increase vaccination uptake by conducting a two-arm cluster randomized study using 10 clinics in the Makwanpur district.

Firat Guder and Tony Cass of Imperial College London in the United Kingdom along with George Mahuku and James Legg at the International Institute of Tropical Agriculture in Tanzania will develop a smartphone-based electrochemical lateral flow assay to rapidly diagnose crop viruses in the field and relay the results. Viral crop diseases like maize lethal necrosis and cassava brown streak disease can destroy up to 30% of crops and are the main threat to food security in East Africa. Most diagnostic tests are laboratory-based and slow, hampering efforts to stop the diseases from spreading. They will develop a system that combines cutting edge chemistry and widely-used smartphone technology to quickly test field samples and upload the results to the cloud for immediate sharing with farmers and agricultural partners. Their approach is based on chemical amplification and detection of nucleic acid aptamers: as few as a hundred virus particles can be detected without the need for complex genetics or fragile and expensive antibodies. They will identify aptamers that selectively recognize the viruses, optimize the lateral flow assay in the lab, and then field-test it in Tanzania and the Democratic Republic of Congo. Once validated, they plan to develop a simpler, disposable version and scale-up the technology to other crops and countries.

Rachel Sklar of Pit Pumpers Ltd. in Rwanda will develop an interactive SMS messaging platform to increase community use of companies to safely empty pit latrines in Rwanda by using communication to increase demand and thereby decrease cost and wait-times for service calls. Pit latrines, used by nearly two billion people worldwide, must be emptied regularly to avoid public health risks associated with fecal contamination of groundwater. Companies in Rwanda provide an emptying service with cost-saving opportunities when several households book together. However, this is more difficult to arrange for users in remote areas, and households may resort to unsafe dumping. They will improve coordination of service requests among remote communities by using local shop owners (duka agents) to identify households needing an emptying service. These leads will be followed up with targeted SMS marketing, and the agents paid a commission for those converted to customers. The interactive SMS messaging platform will provide households with a referral code to invite neighbors to join the service call, and the resulting cluster of customers pay a reduced rate for emptying. It will further include educational messages explaining the negative impacts of manual dumping of fecal sludge to promote behavior change. They will develop the platform and test their approach by training agents in Kigali, Rwanda.

Molly Brown of Syngenta Foundation for Sustainable Agriculture in the U.S. will develop an early warning system for crop diseases for rural farmers in Africa by gathering data using existing infrastructure and mobile tools from commercial partners and applying machine learning pest models. Insect pests cause almost half of crop losses in Africa each year, which impacts both food supply and the economy. An effective disease and pest surveillance system provides early warning to minimize crop loss but is often unavailable to rural farmers in poor countries who lack tools for sufficient pest data collection. To address this, they propose to leverage the existing infrastructure, including demonstration farms, retailers, and mobile tools, of a commercial partner to gather the required data, such as field photographs, data on pest prevalence, and satellite weather data. Mobile tools will be available on computer and Android devices with off-line modes. They will pilot test their approach in Zambia focusing on damage to maize crops caused by the fall armyworm. Real-time data will be uploaded to a freely available, public portal with information on geography, pest populations, and affected crops.

William Martin of the International Food Policy Research Institute in the U.S. will combine existing, high-quality survey data collected from individual households in rural Ethiopia and Nigeria with agricultural information from the Global Agro-Ecological Zone (GAEZ) database to model the impact of specific investments on poor communities to better inform policy decisions. National Agricultural Investment Plans (NAIPS) shape policy development by outlining investments required to stimulate growth in Africa through transformation of the agricultural sector. However, they rarely consider the impact on individual households. To address this, they will use existing household survey data from the Living Standards Measurement Study – Integrated Services on Agriculture (LSMS-ISA), which was designed to dissect the relationship between agriculture and poverty. These data will be combined with the publicly available GAEZ database, which assesses agricultural resources and potential, to model the direct impact of policy decisions, such as investments in research and rural infrastructure (irrigation, electrification), on households. Results of the analysis will help prioritize future agricultural investments for maximum impact on the poorest families.

Federico Costa of the Federal University of Bahia, Brazil, Mitermayer Galvão dos Reis of Fiocruz, Brazil, and Nathan Grubaugh and Albert I Ko of Yale University in the U.S. will establish metagenomic next generation sequencing in clinical settings in an urban region of Brazil classified as an infectious disease ‘hot spot’ to help develop new diagnostics and identify emerging pathogens. Rapid urbanization in Salvador, a metropolitan of 2.9 million inhabitants in Northeast Brazil, has produced a favorable environment for the emergence and spread of infectious diseases, and was the founding site for the recent Zika epidemic. Early detection is critical for preventing disease spread, particularly in Salvador, which is a transport hub and popular holiday destination. However, diagnosis can be challenging in low-resource settings, especially when the causative pathogen is unknown, the disease has diverse symptoms, or a known pathogen starts causing new symptoms. They will collect around 160 clinical samples from patients with suspected infections at a local infectious disease reference hospital and a maternity unit and apply a next generation sequencing approach together with the IDseq analysis platform to identify pathogens from the sequencing data.

James Berkley and Abdi Abdirahman of the University of Oxford in the United Kingdom will test whether metagenomic analysis of clinical samples from patients with suspected infectious diseases can better identify the causative pathogens than current diagnostic methods, to help improve treatment. In Africa and Asia, many severely malnourished infants die after being discharged from a hospital likely due to infectious disease syndromes such as pneumonia, diarrhea, and sepsis. However, the causative pathogen remains unidentified in the vast majority of cases due to the limited sensitivity of current diagnostic methods. Another group highly vulnerable to the limitations of diagnostics currently available in low-resource settings are those with suspected viral encephalitis, which can cause acute seizures and coma. If diagnosed correctly, some of these diseases are treatable. They will use next generation sequencing to perform metagenomic analyses of samples from patients in both groups to identify the causative pathogens, and test whether first isolating extracellular vesicles that may concentrate viral nucleic acids from the samples can improve sensitivity.

Thushan de Silva, Abdul Karim Sesay, Helen Brotherton, and Beate Kampmann of the Medical Research Council in the United Kingdom will locally implement equipment and methods for next generation sequencing of a range of clinical sample types to detect infectious pathogens in hospitalized neonates in low-resource settings. Almost three million children under five years old die each year in sub-Saharan Africa, which is the highest rate globally. A substantial proportion is likely to be caused by pathogenic infections, including multi-drug resistant organisms. However, defining the etiology of neonatal infections, which is crucial for timely and effective treatment and to block disease spread, remains challenging in countries with limited resources. They will test the potential for next generation sequencing to overcome this challenge by conducting a pilot study using clinical samples from around 30 hospitalized neonates within an ongoing clinical trial in a low-resource setting in West Africa. They will optimize protocols on site for detecting pathogens from a range of sample types and develop methods to specifically detect relevant anti-microbial resistant strains. They plan to use the study to install a network of sites across the region that can perform next generation sequencing and share the resulting large datasets with other sites.

Ophelia Venturelli of the University of Wisconsin-Madison in the U.S. will study the growth kinetics and microbial interactions of a synthetic bacterial community in order to optimize bioreactor design and produce large quantities of mixed cultures at low cost. Mixed microbial populations are used to reconstitute the healthy gut microbiota in infants and children who have suffered malnutrition. However, affordable, large-scale production of microbial communities for biotherapeutics is challenging, in part because of poor understanding of how the growth and viability of individual strains are impacted by environmental factors and the presence of other microbes. To address this, they will build a dynamic model of a synthetic 12-member community that mimics the functional and phylogenetic diversity of the human gut microbiome and monitor its stability and growth over time and in response to variations in culturing parameters. This information will be used to predict community growth and stability in bioreactors, which will then be tested. Their approach is low cost, scalable to industrial-sized bioreactors, and can be generalized to other microbial communities relevant for human health.

Samuel Dorevitch of the University of Illinois at Chicago in the U.S. will build solar-powered ozonation systems to supply purified water to families living in Kenyan slums. Many peri-urban informal settlements (slums) around the world lack safe, affordable drinking water. In the absence of centralized water purification, methods like chlorination, solar disinfection, and filtration can be used. However, these are time-consuming and expensive, and are generally not monitored for water quality. They have developed microplasma technology for a solar-powered system that can purify surface water by ozonation without the need for electricity. They will increase the scale of this system to produce 2,000 litres of clean water each day and add sensing and reporting technology for sending information on water quality to local health officials. To test their approach, a purification system run by an array of solar panels will be installed in each of two slums in Kisumu city to provide water for 50 families and be run by local clean water teams. Two frameworks for managing and sustaining the clean water system - a membership-based cooperative and a for-profit vendor charging a nominal fee – will also be tested. The results will be used to inform the next stage of expansion to systems with a 10,000 litre/day capacity in cities in two African and two South Asian countries.

Galaletsang Tsontswane of Congretype in South Africa will design low-cost, solar-powered insect detection traps equipped with wireless sensors to capture images of insects and transmit them to a central control point to improve rural crop surveillance in developing countries. Crop loss due to pest infestation negatively impacts both food supply and local economies, while rural farmers in developing countries lack resources to monitor crop disease and infestations and are unable to respond before substantial loss occurs. They will design and deploy intelligent insect traps equipped with cameras and test their ability to capture images of trapped insects within a study area and to wirelessly transmit the data to a central location via an existing TV white space network. The data can then be stored and accessed by local farmers and agricultural extension workers, allowing them to respond more quickly to crop threats than is currently possible based on data from laborious, human-based monitoring systems.

Fatema Khatun of the International Centre for Diarrhoeal Disease Research, Bangladesh in Bangladesh will develop a digital intervention to enable sharing of existing digital health data between community health workers and provide them with feedback indicators along with tailored messaging to parents to improve timeliness and coverage of vaccination against tuberculosis in rural Bangladesh. Tuberculosis is the number one cause of death by infectious disease worldwide, and 95% of deaths occur in developing countries. In Bangladesh, the bacille Calmette-Guerin (BCG) vaccine against childhood tuberculosis is usually administered at six weeks of age, despite the fact that vaccination at birth could decrease mortality by three percent – a significant number of cases in this tuberculosis-endemic country. Immediate vaccination is especially important for high risk neonates including low-birth-weight babies and those born to hepatitis B-positive mothers. The digital intervention strategy will be developed to combine the existing electronic health record system with the pregnancy and childbirth registry and to provide digital communication between the community workers providing prenatal and delivery care and those providing BCG vaccination. This will ensure that newborns at increased risk for tuberculosis are prioritized for immediate vaccination. The system will also send direct messages to mothers before and after birth, reminding them of the importance of timely vaccination. Their approach uses existing data in a new way to improve overall vaccination rate and timeliness and decrease infant mortality.

Matthew Edmundson of Violet Health in the U.S. will develop iron-rich biscuits and tailor marketing campaigns to combat iron deficiency in adolescent girls in India. Iron deficiency is a global health concern and is particularly dangerous during pregnancy when it can increase the risk of maternal death and health problems for the infant. Nearly half of all adolescent girls in India are iron-deficient, and although iron tablets are available they are not taken properly, partly due to their bad taste and a cultural aversion to tablets. Thus, more culturally acceptable alternatives are needed. To address this, they developed a low-cost, iron-rich biscuit that could overcome anemia and non-compliance to iron tablets in clinical tests with pregnant women in India. They will now focus on helping low-income adolescent girls by adapting the biscuits to their nutritional needs and preferences, which will be determined by interviewing 50 girls from different areas. These insights will also be used for a pilot marketing campaign to generate demand amongst the girls and their families and community members. They will test their approach with 300 girls from rural and urban locations in India to determine the effects of different marketing methods on demand.

Lyndon Paul of Vissot Co Ltd in Cambodia will reduce production costs for their nutritional wafer biscuits, which are made from a micronutrient-fortified fish powder, to help treat severe acute malnutrition in children and prevent malnutrition in young children and pregnant women in Cambodia. Acute and chronic malnutrition are a major public health concern in Cambodia. They previously developed a fortified fish powder and showed that it could replace milk in food for infants and was effective at reducing malnutrition. However, unstable supply and variable quality of the inland fish used to make the wafers have led to fluctuating prices. To address this, they will set up an optimized supply chain to reduce production costs by 60%. They will train workers in five communities where the fish are caught to sort, clean and pack the fish for transport to their factory in Phnom Penh. There, the fish will be processed into fish meal with acceptable taste and texture. They will evaluate the supply chain by collecting data from the fishers to the final product and evaluate quality and food safety.

Muthupandian Ashokkumar at the University of Melbourne in Australia, along with Francesca Cavalieri, Meifang Zhou, and Srinvas Mettu, will produce edible microballoons made from protein that contain essential nutrients for adding to common foods to combat malnutrition in mothers and infants. Encapsulating the nutrients, rather than adding them directly to food, helps keep them stable and promotes their absorption in the body. It can also mask unpleasant tastes, and control the timing and location of nutrient release, which can increase their performance. They have developed a method that uses ultrasound waves to encapsulate oil- and water-soluble vitamins and minerals within edible shells made from a range of proteins including milk and pea proteins. They will analyze the stability and strength of microballoons made from different materials that contain the recommended daily doses of nutrients for mothers and infants. They will also develop methods to encapsulate water, which could be used to reduce the fat content of fat-rich products.

Gloraia Pena of Cooperativa Multiactiva De Madres Del Valle Coomac in Colombia will implement a hybrid value chain business model to leverage collective purchasing power in a community of low-income families in Colombia to reduce the price of nutritious local foods. Current food prices are relatively high for low-income families because they buy in small volumes. They will combine collective purchasing power with a hybrid value chain model, which incorporates the needs and roles of the public and private sectors, to increase access to nutritional foods. They will collect social and economic data from an existing group of 9,000 families in a poor neighbourhood in Colombia to understand how their approach should be implemented. This will include the numbers of participants needed to reduce the cost sufficiently to encourage people to buy the healthier foods and ultimately produce a positive long-term impact.

Amanda Stiles of Ripple Foods, PBC, in the U.S. will produce a low-cost protein isolate upcycled from locally-sourced agricultural by-products that can be used as a nutritious food additive or standalone high-protein broth. Protein malnutrition is a major health concern in southern Asia and sub-Saharan Africa. However, protein is expensive to produce and often has a bad taste. They have developed an automated approach to identify low-cost, efficient methods to isolate plant proteins from agricultural by-products in the U.S. They will apply their approach to by-products from low-resource settings, such as wheat bran, and perform a high-throughput protein isolation screen to identify optimal extraction and purification protocols for yield and purity. The final products will be taste-tested to ensure they have a limited impact on flavor when used as food additives.

Joshua Warren and Daniel Weinberger of Yale University in the U.S. will develop an analytical framework to improve local estimates of vaccine coverage in low- to middle-income countries. Current estimates can be unreliable, due to errors and biases in record-keeping and difficulties in estimating local population sizes, and are further complicated when children are vaccinated outside of their home administrative district. They will develop advanced spatial analytical methods including bias adjustments that take these issues into account and can generate more reliable local estimates also from poor quality data. They will collect higher quality survey data on vaccine coverage and population sizes from selected locations to calibrate and ultimately validate the estimates. Their approach can be used in other low-income settings to improve vaccine coverage.

Ali Turab of IRD Global Ltd. in Singapore will develop a decision support tool that can be integrated with digital immunization registries to automatically construct optimal appointment schedules for every child that can adjust for missed immunizations and the introduction of new vaccines. A large majority of children, in both developing and developed countries, are not immunized at the recommended times, which can increase the risk of severe diseases. When a vaccination is missed, it is left to the health care professional to work out the best alternative schedule, which is often inaccurate. To help with this, they will design software that incorporates a child's vaccination history and age to automatically construct a new immunization schedule after every appointment, and that can also identify opportunities to vaccinate children even when they are at a clinic for other reasons. The software will integrate with existing health systems in several developing countries. They will conduct a mixed methods study at the Indus Hospital Korangi Campus in Pakistan to validate their approach for generating optimal schedules and assess usability by health workers.

James Njeru of the Field Epidemiology Society of Kenya will develop an integrated electronic platform that collects immunization and health data from existing registries and automatically sends regular, user-defined reports via SMS and email to health workers to improve vaccine coverage. Healthcare facilities record their immunization data on District Health Information Systems, but access to the data is limited. To widen access, they will build a platform that analyzes relevant health data from various sources, which will improve data accuracy, displays it on dashboards, registers users, and tracks their activity. They will pilot test their platform over seven months by registering program staff and managers so that they can access the platform and receive reports. The platform will be evaluated for its ability to integrate data and produce reports such as coverage and dropout rates. Feedback from users will also be used to refine the platform.

Chibuzo Opara of DrugStoc E Hub Ltd. in Nigeria will equip vaccine storage and transport sites with calibrated weighing mats (Digimats) that automatically transmit vaccine quantities in real time to better monitor delivery chains in the community and improve supply. Monitoring the movement of vaccines at the national and district level is currently performed by the Nigerian immunization program. However, accurate monitoring at the local level requires alternative, more automated approaches to avoid human error. They will calibrate their Digimats to recognize the weight of specific vaccines, and identify 20 sites across three states, including storage warehouses and trucks, where they will be positioned to automatically transmit data over a period of six months. These data will be collected by mobile tablets and interfaced with the national vaccine delivery dashboard to provide real-time stock counts and resupply alerts.

Mustafa Naseem of the University of Michigan in the U.S. will apply machine-learning algorithms to identify potentially falsified digital vaccination records in Pakistan. Pakistan is one of only three remaining countries where polio is still endemic. Particularly rural healthcare facilities are struggling to provide enough vaccinations due to highly populous provinces and a lack of resources and staff, and there is a risk that records are falsified to save time or bias the results. They will first perform fieldwork to identify any putative recently falsified records by auditing 2,000 recorded vaccination events across 200 randomly-selected villages. These data will be used to generate an algorithm by using features such as record patterns that can then detect if a data-point is likely to be true or false. They will test their approach by auditing another 1,000 vaccination events that the algorithm predicted were falsified compared to 1,000 randomly selected vaccinations.

Julius Lucks of Northwestern University in the U.S. will develop a low-cost diagnostic test that can detect multiple plant pathogens with a simple visual output for farmers in low-income countries to better monitor their crops. Current diagnostic field tests can only detect one disease and are generally difficult to use and costly. The alternative is laboratory testing, which is often unavailable in low-resource settings. Taking a different approach, they will develop a sensitive, multiplexed test that only requires basic sample preparation, such as mixing and using body heat, and can detect multiple pathogens using biosensors. The results will be visually presented using color changes that can be recorded by cell-phone cameras for analysis and reporting to aid global plant pathogen surveillance efforts. They will develop the methods and tools to detect three model plant pathogens and field test their diagnostic system in the U.S., Uganda, and Kenya.

Jun Kameoka of Texas A&M University in the U.S. will develop multiplex, battery-less and wireless durable paper sensors for positioning under the soil in crop fields to detect the early signs of pests and diseases, and communicate the data to overhead drones via radio frequency to improve pest management. The sensor will be designed to monitor physical, biological and chemical soil conditions that are altered by plant diseases. They will test its performance in commercial garden soil with maize and sorghum plants in a vinyl house.

Jan Kreuze of the International Potato Center in Peru will develop a low-cost, mobile phone-based diagnostic test for African farmers that uses artificial intelligence to quickly and accurately detect plant diseases such as cassava brown streak and banana bunchy top, which devastate crops and are threatening to spread. Accurately diagnosing plant diseases is difficult because visual symptoms can be highly variable. Artificial intelligence (AI) has shown promise for analyzing images of plants taken by mobile phone to detect diseases in low-resource settings, but it is not accurate enough. Alternatively, chemical-based diagnostic tests that detect the underlying viruses are far more accurate but difficult to use without training and require costly equipment. They will enhance the accuracy of AI for diagnosing a range of plant diseases by mobile phone by training it with validated diagnostic test results from their microfluidic amplification and detection device used by researchers and inspection agents. Their approach has the potential to recognize hard-to-detect symptoms in plants that may even be missed by crop specialists.

David Hughes, and Nita Bharti of Penn State University in the U.S. together with James Legg at the International Institute of Tropical Agriculture in Tanzania and the Charity, Self Help Africa, will leverage daily, high-resolution satellite imagery of farms in Kenya to monitor crop pests and diseases. Publicly funded satellites have the capacity to measure crop health, soil moisture, and water availability across wide areas. However, they are unable to accurately diagnose crop diseases particularly in smallholder farms because of the presence of many different types of often unhealthy-looking vegetation caused by lack of water or nutrients rather than plant diseases. They will use ground data on crop diseases and pests being collected as part of a five-year EU-funded project at 1,400 farms in seven counties growing a variety of crops. They will also collect maps of the farms using drones flying at different heights and see how well any pests and diseases can be detected using the daily satellite data. They will validate their approach for detecting pests and diseases on an additional 1,400 farms.

Matteo Rinaldi of Northeastern University in the U.S. will develop a miniaturized, maintenance-free chemical sensor that can detect specific volatile organic chemical vapors released from diseased crops as an effective surveillance system suitable for low-resource settings. Manual surveillance is time-consuming and requires prior knowledge of disease symptoms. Automated, sensor-based crop surveillance is far more effective, but relatively expensive, and the sensors constantly consume power, making them unsuitable for low-resource settings. They will develop a low-energy sensor-based monitoring system by exploiting a recently developed technology that comprises a micromechanical switch made of two cantilever beams. One of the beams will be coated with a polymer sensitive to the plant-based chemical and exposed to the environment. In the presence of that chemical, the beam undergoes a change in mechanical stress, causing it to bend and make contact with the second beam to trigger the switch. They will develop the microswitch-based chemical sensors, integrate them with a low-power long-range wireless module to signal pest detection, and test the performance of prototypes in the laboratory.

Monica Nolan of MU-JHU Care Limited in Uganda will adapt the existing open source Smart Register Platform, which digitally stores health records, for the real-time collection and transfer of immunization data, to improve vaccine coverage and other healthcare services for women and children in Uganda. In many low- to middle-income countries, records of childhood vaccinations are usually written by hand and can be poor quality. Digital records are of better quality and value, as they also enable the integration of different types of healthcare services, such as HIV services and vaccinations, to improve overall health. They will adapt existing technology and infrastructure, including the Smart Register Platform, which is already integrated into several national health systems and can produce automated SMS reminders of appointments. They will also design methods informed by mothers with young families, health workers and managers, to optimize data use and delivery of health services. They will use surveys and analyze health data to evaluate their approach for improving vaccine timeliness and coverage at selected clinics.

Chinedu Chugbo of Avigo Health L.L.C. in the U.S. will develop an approach to crowdsource reports of infant births and deaths from community members by health workers to better monitor vaccine coverage in low- to middle-income countries. In Nigeria, only 30% of births are registered, making it difficult to estimate numbers of vaccine-eligible children and ensure every child is properly vaccinated. Current methods for estimating population sizes include household surveys, which are costly, or records from health clinics, which suffer from limited coverage. Crowdsourcing is a proven method for efficient data collection, although data quality may be variable. They will develop electronic data-collection and storage tools and pilot test their crowdsourcing approach in a selected region in Nigeria. Health workers will be trained to administer brief interviews to community members visiting clinics and during outreach programs to document local births and deaths. They will evaluate the performance of their approach and particularly data accuracy by comparing it with data collected by household surveys in the same region.

Brandon Kohrt of George Washington University in the U.S. will develop a tool using sensors associated with mobile phones that can identify and monitor young mothers suffering from perinatal depression in low-resource settings in Nepal so that personalized psychological treatments can be provided. The tool comprises a mobile phone for the mother and a small Bluetooth beacon attached to the baby's clothes. It can record location via GPS, the proximity between phone and beacon, and sound. They will build sensor data models that can associate specific activity recorded from the sensors with maternal depression, such as the length of time spent with the baby, the level of vocal interactions, and the frequency of outings. A community advisory board composed of young and older mothers, family members, and health workers will be established to evaluate the approach and help test and refine the tools. They will also co-develop a user-friendly interface with the mothers and health workers so that they can easily access the information and use it to improve their mental health.

Allya Paramita Koesoema of the eHealth and Telemedicine Society in Indonesia will develop a pro-vaccination campaign to counteract the widespread negative views of vaccinations in Indonesia by engaging religious leaders and health workers in local communities to directly address misconceptions. Anti-vaccination narratives, many based on religious misconceptions, have spread through the country, largely via social media, leading to a decrease in child vaccination coverage. Health workers often do not have the knowledge to explain away these misconceptions to mothers when they refuse to have their children vaccinated. To address this, they will engage stakeholders to identify the anti-vaccination messages and build a database of effective reasoning to directly challenge those messages that can be accessed by health workers and other respected, senior community members. They will perform a randomized controlled trial in different districts in Indonesia, and supplement the pro-vaccination messengers with vaccination reminders and schedules sent directly to mothers via mobile phone. They will test the usability of their approach and its effect on the willingness of the mothers to vaccinate.