How Satellite Data Helped Get Food to the Hungry during COVID-19

A COVID-19 emergency food program in the Philippines offered an opportunity to design a targeting program based on granular poverty maps that were compiled using traditional and innovative data sources and artificial intelligence. Photo credit: ADB.

Share on:           


Poverty maps derived from satellite images helped target the most vulnerable households in pandemic-affected areas in the Philippines.


Every day, the world generates an estimated 2.5 quintillion bytes of data. Applications of data come from digital transactions, telecommunications records, social media, remote sensing, to name a few, which permeate almost every aspect of daily life.

Yet, when the coronavirus disease (COVID-19) struck, everyone was blindsided by the lack of information on the novel virus. Data have since been collected and analyzed, providing actionable insights on how to navigate through this crisis. 

As countries rebuild their economies, it is important to maximize the use of data to ensure that nobody, especially the poor, will not be left behind. Before COVID-19, it was estimated that the proportion of people living in extreme poverty in developing Asia would have declined to 114 million by the end of this year. Because of the pandemic, the region’s poorest will likely increase to around 192 million instead.

But how can the much-needed assistance be given to the poor if the answer to a simple question, like where the poor are, is not as straightforward as it seems? Poverty statistics are usually compiled by conducting a household income or expenditure survey or a living standards survey. However, survey sample sizes are typically not large enough to pinpoint where the poor are at a granular level, and increasing sample sizes can be costly.

Research shows integrating conventional and innovative data sources, such as satellite images, offer a cost-effective approach to collecting granular poverty data. A COVID-19 emergency food program in the Philippines offered an opportunity to raise this approach from a mere academic exercise to a real-world application.

Poverty Targeting in an Emergency

Bayan Bayanihan showcases how novel applications of big data and technological tools can inform the design of programs that benefit the poor. The program distributed critical food supplies to the poorest areas of Metro Manila during the onset of the lockdown to control the spread of the virus.

When the Enhanced Community Quarantine (which restricted the movement of the population to work, health-related, and other essential activities)  started in Metro Manila in mid-March, the Asian Development Bank (ADB) acted swiftly by launching a food program in partnership with the Philippines’ Department of Social Welfare and Development and the private sector and in coordination with the Philippine Army. The name “Bayan Bayanihan” is based on the traditional Filipino bayanihan principle of community spirit. In a span of 2 months, the program served basic food commodities (e.g., rice, canned goods) to 162,000 households across 44 barangays not just in the metropolis but also in nearby provinces.

Data played an important role in identifying the food program’s target beneficiaries to ensure that the poorest and vulnerable areas were prioritized and provided immediate assistance. The targeting program relied on granular poverty maps compiled by ADB statisticians, using a combination of traditional and innovative data sources and artificial intelligence.

The poverty maps were produced by training a computer vision algorithm to spot specific features from daytime satellite images to predict the level of economic activity in an area. An earlier initiative trained the algorithm to use the intensity of night lights initially as a proxy for economic development and made adjustments to estimate poverty in the Philippines. Since satellite imagery is available for granular areas, this method can produce poverty maps at granular levels too, making targeting more efficient.

Design Considerations

One of the main considerations in identifying priority barangays for the program is the prevalence of poverty in the area. The project team looked at above-average poverty levels, size of the population, and distance from shops to identify the most vulnerable households.

Data pertaining to the presence of retail facilities and markets in a barangay were taken from the Census of Population and Housing of the Philippine Statistics Authority. By integrating this data source with the poverty maps, the program could prioritize areas that were poor and whose residents may encounter more difficulties in accessing food due to longer distances to markets.

The two data sources were further triangulated with real-time information coming from the field. For example, during the program’s implementation, there were referrals made on the ground to consider other poor areas. Some of these areas were included in the list of beneficiaries after field validation.

The project team also worked with the Philippine Army to check the logistics of distribution and to draw up detailed logistics plans to organize how many trucks and people and goods to be distributed per day.

The Importance of High-Quality Data

The food relief provided through the Bayan Bayanihan program concluded in May 2020, but ADB continues to support the Philippines and other developing members in responding to the COVID-19 crisis through finance, knowledge, and partnerships. 

The social and economic shock waves caused by COVID-19 continue to spread across the globe, and its cumulative effects are expected to be felt for many years to come. It is important to formulate a data-guided set of actions to protect the poor, the vulnerable, and wider populations across Asia and the Pacific.  To accomplish this, the use of targeting mechanisms driven by high-quality data will play a critical role to ensure that a program serves those who are in most need while using resources efficiently.

The Bayan Bayanihan program shows how alternative sources of data (such as those coming from satellite images) can be integrated with conventional survey data to generate granular poverty data for policy targeting purposes. Other countries that plan to do the same program would need to follow a similar method for data analysis and for distribution logistics.

Arturo Martinez, Jr.
Statistician, Economic Research and Development Impact Department, Asian Development Bank

Art Martinez works on Sustainable Development Goals indicator compilation, particularly poverty statistics and big data analytics. Prior to joining ADB, he was a research fellow at the University of Queensland where he also got his doctorate in Social Statistics.

Anouj Mehta
Country Director, Thailand Resident Mission, Asian Development Bank

Anouj Mehta leads the planning, implementation, and supervision of Thailand Resident Mission's vision, goals, strategies, and work plan. Before his current role, he led the pioneering ASEAN Catalytic Green Finance (ACGF) facility under the ASEAN Infrastructure Fund. He also set up and managed the Innovation Hub for the Southeast Asia Department, and led one of ADB’s pioneering public–private partnership initiatives in India. Prior to joining ADB, he was an investment banker and chartered accountant at JP Morgan Chase and PWC in London.

Follow Anouj Mehta on

Asian Development Bank (ADB)

The Asian Development Bank is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 68 members—49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance.

Follow Asian Development Bank (ADB) on
Leave your question or comment in the section below:

The views expressed on this website are those of the authors and do not necessarily reflect the views and policies of the Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. By making any designation of or reference to a particular territory or geographic area, or by using the term “country” in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area.