Using Artificial Intelligence (AI) and Machine Learning (ML) techniques to identify and quantify populations in rapid growth informal areas.
Identify rapidly growing populations in urban informal areas, so as to better target populations in need of immunization for GAVI partners.
Describe what you intend to do and how you'll do it in one to two sentences (required 350 Characters)
iMMAP aims to use innovative AI and ML techniques using high resolution satellite imagery to identify rapidly growing populations in urban informal communities in two cities in Yemen. To do this work we will partner with DigitalGlobe, a leading commercial satellite company with experience in applying AI and ML techniques from space.
Explain the innovation (2,500 characters)
Besides the usual global urbanization trends, conflict areas are especially seeing unprecedented urbanization as rural populations flock to urban centers due to rural livelihoods being destroyed by conflict. Populations are forced to move to try and gain access to basic services. As a consequence these rapidly growing areas are often missed in immunization activities due to the lack of updated data on rapid informal growth areas, and numbers of people. This methodology aims to provide up-to-date information on where rapid growth urban areas are, and also provide dis-aggregated data on estimated population figures by age and gender. iMMAP and DigitalGlobe propose a rapid mapping of informal areas / slums in cities, while also providing a change map over the last 5 years to identify fast growing informal areas. We will use a combined approach of feature mapping from DigitalGlobe EarthWatch with crowd-sourcing (Geohive) for feature identification and/or AI & machine learning to identify slums in cities. Running a change detection map over the last 5 years, based on the availability of archive, would be definitely a great asset. Population mapping, be it counting houses, accommodations, or even tents, will be conducted with a scalable approach using AI / Machine Learning / Crowd Sourcing. Leveraging DigitalGlobe’s archive of data over the proposed area of interest is of high interest to be able to detect the changes of population settlements. By identifying where these informal communities are, and quantifying the population, GAVI partners will benefit from better immunization targeting of populations that are often neglected. Answering questions such as: Where do they live? What is their age? How many people might this intervention realistically reach? will be especially useful for on-the-ground immunization teams, while also creating a city level evidence base for GAVI on where new pockets of vulnerable people are.
Which part(s) of the world does this innovation target?
For a pilot phase, iMMAP proposes to focus on two cities in Yemen (Sana'a and Aden), specifically targeting urban informal communities. Once proven successful, the methodology can be expanded to other GAVI countries.
Who will work alongside your organization in the project idea? (1,000 characters)
iMMAP will partner with DigitalGlobe, a commercial satellite company. DigitalGlobe owns and operates the most agile and sophisticated constellation of high-resolution commercial Earth imaging satellites. They have vast experience in using AI and ML techniques for applications such as: Eliminating Malaria from Space, eradicating Polio, Population Density from high resolution satellite imagery, Ebola Response, Migration and Famine, Monitoring Refugee camps / Refugee migration, and mapping poverty from space. DigitalGlobe also has a crowd-sourcing platform to enable rapid mapping using volunteers.
Globally iMMAP has partnership agreements with nine United Nations agencies, and several International NGOs, and we have been partnering with DigitalGlobe for the last two years using their EarthWatch platform for up-to-date satellite imagery in urban areas.
How is your idea unique? (750 characters)
The proposed project to identify informal settlements using satellite imagery is unique in its method as it uses AI algorithms to identify living neighbourhoods and any sign of inhabited place. The crowd-sourcing platform allows then to qualify the dataset and eventually create learning dataset for AI to perform better. In the area of conflicts, deserted or still populated places can be hard to detect and sometimes the human eye is the key differentiator. The strength of the project is to ally human inputs where machine still have poor results, but to heavily use multiple high end technologies hosted by DigitalGlobe with no requirement of bespoke IT hardware. It represents a great result per investment ratio, avoiding purchase data that wil
What is the name of your organization
Explain your organization (250 characters)
We support partners to solve operational and strategic challenges. Our pioneering approach facilitates informed and effective emergency preparedness, humanitarian response, and development aid activities by enabling evidence-based decision making.
Type of Submitter
We are a registered NGO or Non-Profit Organization
Gender and Diversity (500 characters)
iMMAP is an equal opportunity employer and the recruitment process is conducted without discrimination of any kind based on race, colour, sex, religion, political or other opinion, national or social origin, birth or other status such as disability, marital and family status, sexual orientation and gender identity, economic and social situation.
iMMAP follows the Inter-Agency Standing Committee (IASC) guidelines on gender and inclusion of people with disabilities in data collection efforts.
Organization Location (less than 250 Characters)
iMMAP has global headquarters in Washington, DC, and France. We have a representation office in Geneva, and Regional offices in Amman, Bangkok and Panama. We have strong country presence in Nigeria, Ethiopia, Yemen, Afghanistan, Colombia, and Iraq.
Size of organization (number of employees):
Scale of organizational work
Global (within 2 or more global regions)
Applying to Gavi INFUSE
Referred by a friend / colleague