‘Biometrics are not a magic solution’ Simprints CEO cautions
Using biometrics for the sake of it can jeopardize projects, according to the CEO of biometrics firm Simprints. In his presentation to the Global Digital Development Summit, Toby Norman calls on those in the development sector to scrutinize any desires to incorporate a biometric element to their initiatives.
While there is more data available on specific elements of projects than ever before, Norman believes more is needed on the broader delivery of services. But biometrics are not always the answer and sometimes far more simple metrics are far more effective to evaluate success.
Simprints, a non-profit, helps organizations delivering projects such as healthcare on the frontline. It finds that in the right situations, biometrics can allow huge improvements in service delivery, such as a project that saw a 38 percent increase in women seeking antenatal care in Daka, Bangladesh.
But CEO Norman reports problems he has had on his own projects as a PhD researcher, prior to the founding of Simprints, where follow-up survey data does not match provision data, which once led to the dumping of medicines in storage. Vaccination programs could suffer the same gaps between the administrative data collected by frontline workers and any academic estimates or surveys, which only provide a snapshot.
Digital health service records can help, but the lack of foundational or functional ID systems can mean this data is not always available. Introducing biometrics for a specific project can help solve the ‘who?’ question, according to Norman, but it is not always the best approach.
“Every time that we introduce a new intervention and a new technology, we have to be really thoughtful about where and how we deploy,” says Toby Norman. “Biometrics are not a magic solution to digital programs. It the project is a giant mess, biometrics can help support it becoming a digitized, well-monitored mess, but it’s not going to change the fundamentals.”
He lists a set of scenarios where biometrics can help a project. For example, when the intervention has monetary value – such as a vaccine; where incentives exist for misreporting progress; longitudinal projects such as when recipients need to be given a certain treatment of a given time period, typically illnesses including TB and HIV; projects where the target population has few alternatives; projects that will be measured digitally and where there is no other way to report.
Norman gives the example of distributing mosquito nets. Instead of a costly and complex rollout of biometrics, a simple photo of a bed net in place would be far more effective.
Giving another example of biometrics, as firms shift to contactless, he cites a Simprints study which found around 30 percent of women’s fingerprints changing significantly over a three-year period due to cooking burns.
Norman urges those working in development to answer four questions before embarking on adding biometrics to an initiative: Are biometrics necessary? Will they solve the problem? Which tool is appropriate? How will privacy be protected?
The CEO encourages the use of GDPR privacy legislation as he believes it to be robust and to work well with biometrics.
Looking to the future as biometric use becomes increasingly common, the co-founder hopes more APIs will bring costs down as integrations go up. He hopes firms supplying biometrics and software will move away from per-transaction fees and that there will be greater interoperability across platforms.