Infant Biometric Identity for Health & Civil Service

Mobile delivery of biometric identity to link the world's youngest populations to essential health and civil services.     

Of the 1.1 billion people worldwide who do not have formal identification, the majority are children living in Africa and Asia. While countries have adopted digital technologies like biometric identification to address their identity challenges, there is no commercially-available biometric solution for infants and young children - for whom privacy and control over personal data is perhaps most important. As healthcare, civil, and social protection services are increasingly delivered digitally, the lack of appropriate technologies for this age group could perpetuate their exclusion.

Element specializes in mobile delivery of biometric identity. Our software-only solution for biometric recognition uses the built-in cameras on mobile devices - no specialized hardware or connectivity required. We apply Deep Learning - a modern form of Artificial Intelligence (AI) - techniques to multiple modalities to build a contactless, portable platform designed to scale across the 3 billion smartphones that already proliferate the world. It’s a future-proof, cost-effective, and natively-secure approach (the abstractions generated by the models cannot be reverse-engineered into the original images) that is designed to integrate into health and civil systems.

Over the last two years, we’ve been working with the Global Good Fund and partners in Bangladesh, Cambodia, and Mozambique to extend the platform to infants and young children - creating the first biometric platform capable of following a person from birth through old age. We conducted the world’s largest infant biometric trial, training our models on biometrics images of palms, ears, and feet from 8,000 infants and children, and developed a mobile prototype currently undergoing field validation on 2,000 infants. We have developed a three year strategy to scale the solution to 100,000 infants and young children to deliver a globally available solution.

This roadmap has attracted a world-class group of partners. We were selected by Gavi, the Vaccine Alliance to scale the platform in support of immunization and are engaging with UNICEF to inform considerations for appropriate use of biometric technology on children, including data security and privacy. We are also exploring collaborations with the Technical Support Unit for CRVS in Pakistan. 

As leaders of the Principles on Identification for Sustainable Development, ID4D is the ideal partner to join and co-lead this work. Through Mission Billion, our hope is to engage ID4D to ensure the infant biometric solution is built to fit the needs of digital identification systems in developing countries and that the right frameworks on privacy and data control are in place to safely foster its uptake. By delivering a safe, scalable, and sustainable digital identity solution for under-five populations, we’re looking to empower the youngest beneficiaries in accessing health and civil services.

There is no commercial solution for infant biometric recognition. Early-stage hardware fingerprint prototypes have met challenges such as infant behavior, accuracy, and distribution models capable of reaching last-mile populations. Element’s mobile software solution was built to solve this combination of performance and delivery requirements.

Element is a global leader in state-of-the-art implementation of Deep Learning algorithms local to the mobile device for secure recognition. We apply these techniques to multiple modalities which are robust, protect privacy, and are easy to capture with typical infant behavior. The lightweight architecture allows offline enrollment and authentication - essential for scaling in low-connectivity environments.

Element’s solution was designed with native privacy and security features. Using the built-in camera on mobile devices, images of a person’s biometric modality are captured and Deep Learning algorithms create an abstracted “user model” that is unique to the individual and cannot be reverse-engineered into the original images. Biometric images are the only data required for enrollment and are deleted after user model creation - no other PII is necessary. The biometric data is also stored separately from sensitive program data (e.g., health or financial information).

User models are the only data that sync across servers or mobile devices and alone are not sufficient for matching - additional components must be brought together in a secure channel local to an authorized mobile device, at the point of contact with the infant or child, which is usually supervised in a health or civil context. This decentralized approach gives people full control over their data as the person must be physically present to be matched and the data architecture means the user's identity is effectively anonymized outside of local matching. 

Element’s platform creates a digital identity layer that can be accessed across mobile devices – giving users control over their data and helping connect siloed systems in a privacy-centric approach. For example, in a healthcare context, a woman could enroll in a community health program, use the same biometric identity to access a record at a health clinic, and the same biometric identity to verify an insurance claim - without the need for expensive, specialized hardware at point of care or the need to share sensitive information across those platforms. It’s a cost-effective, privacy-centric approach that puts the person at the center of service delivery, which is increasingly mobile-first.

By extending our adult solution to under-five populations, Element aims to create the first identity platform capable of biometrically following a person from birth through old age, matching health and civil service delivery across the lifespan. We would welcome exploring with ID4D how models of self-sovereign control of data independent of civil authority could also be implemented in the context of foundational identity platforms. 

Element is purposely built as a modular identity solution. The technology can be delivered as a standalone mobile application or as a Software Development Kit for integration into existing platforms. We already have integrations with platforms like SAP and DHIS.

The benefit of a software-only solution is that it can be entirely user-centered. Using Element looks and feels like taking a photo – a comfortable gesture thanks to the global proliferation of mobile devices. This makes it easy for government or health workers to use, familiar to the people they work with, and adaptable for a range of social contexts.

Enrollment in Element creates a portable digital identity that can be used to facilitate linkages between disparate systems. We leverage partners’ APIs, follow best practices on data management, privacy, and accuracy, and can facilitate groundtruth storage to prevent vendor lock-in.

Element is committed to maximizing uptake of the solution in low- and middle-income countries, evidenced by our partnerships with Gavi, the Vaccine Alliance and the open source DHIS platform. We are exploring launching the infant biometric platform in a free-to-use basic tier.

Element has spent the last six years focused on mobile software development. As such, our infant solution immediately benefits from the company’s innovations in security and mobile delivery, including ongoing work to optimize for low-end devices and areas of low-connectivity.

To develop the infant biometric platform, we worked with healthcare partners in Cambodia, Bangladesh, and Mozambique. We incorporated end-user feedback directly into our product development cycle - for example, limiting written text and enlarging buttons to be sensitive to varying levels of written and digital literacy, ensuring offline enrollment/authentication, and mapping syncing protocols to data needs (e.g., wifi-only). 

Element has developed a three year strategy to scale the infant biometric solution to 100,000 infants and young children across Africa and Asia. This is designed to contribute to best practices for delivery in low-resource settings, provide sufficiently diverse data to accelerate R&D, and deliver a globally available platform.

By securing the right resourcing and partnerships, Element is exploring launching the infant biometric platform for all the world’s populations in a free-to-use basic tier. This would be supported by a sustainable growth model through commercial deployments of our adult solution.

Element and the Global Good Fund carried out the world’s largest infant biometric trial with Angkor Hospital for Children in Cambodia and ICDDR,B in Bangladesh. Field validation is underway with the Manhiça Health Research Centre in Mozambique. 

Our partners include governments and multilaterals. Element was selected by Gavi, the Vaccine Alliance to scale the infant biometric solution to support immunization delivery. We’re engaging with UNICEF to inform considerations for appropriate use of biometric technology on children, including data security and privacy. We’re also exploring opportunities to implement the solution with the Technical Support Unit for CRVS in Pakistan. 

Element was one of the first modern AI research labs and the first focused on digital identity. With Yann LeCun - a pioneer of Deep Learning - as our co-founder, we attract leading AI researchers and engineers. We require a doctorate degree or master’s degree with equivalent practical experience and specialized experience with Deep Learning frameworks.

Our team is fortified with experts in global health programming, policymaking, and building partnerships to scale innovation. Other initiatives leads come from organizations like Google, eBay, and J.P. Morgan, with diverse expertise in finance, product development, and product marketing.

A software license based on an annual subscription paid per user account per year, or a fee structure based on volume and value of financial transactions facilitated by the service. Element’s partners host the software/data, which enables us to maintain sufficient margins on our software license revenue stream to allow us to continually invest in building and delivering technology advancements over time.

Element operates in over ten countries and works with partners to use mobile biometric identity to link people to services. The under-five platform can be brought to market through this significant, existing channel of partners and multi-country operations.

By securing the right resourcing and partnerships, Element is exploring launching the infant biometric platform in a free-to-use basic tier. The platform is integrated into our core pipeline, ensuring the full support of Element’s research and customer service.

Co-creation with implementers and policy experts is essential to ensure that the infant biometric solution is responsibly built for purpose and that the right frameworks are in place to protect the privacy and rights of children. That’s why ID4D is the ideal partner. Through Mission Billion, we hope to leverage ID4D’s expertise in setting global principles for inclusive and secure identification systems and fostering cooperation around practical solutions. It’s an opportunity for ID4D to help ensure the infant biometric solution fits the needs of digital identification systems in developing countries - including considerations around privacy, security, and interoperability. 

As a Deep Learning powered platform, it is essential that our models are trained on diverse datasets that are collected longitudinally and distributed evenly across age categories.

Pure data collection is not sufficient to deliver a fit-for-purpose solution – it must be accessed via implementation to ensure it is robust to real world conditions. For example, collecting biometric data on the same timeline as an immunization schedule will ensure the models are trained on sufficient data at each immunization milestone age (e.g., birth, 14 weeks, 36 weeks). This is why implementation research is part of our scale-up strategy.