Based on my experience, most discussions about the authentication of individuals using a biometric modality (such as fingerprints, or voice or facial recognition) often just focus on key issues such as reliability, security, ease of use, cost, and privacy concerns. Certainly these are important issues, but one that is often omitted in the conversation is the use of a biometrics system for health and safety purposes.
My wife and I were recently blessed with the birth of our fifth grandchild, a beautiful baby girl. During the hospital visit, the risk management side of me evaluated the security aspects of the facility. What methods prevent the accidental swapping of babies or the theft of a newborn? While the frequency of such incidents in developed countries is very low, it is a more challenging issue in developing countries where medical recordkeeping is often minimal and limited to paper documents.
Talking to the hospital staff, I found out they have a number of safeguards in place to ensure the right baby is with the right mother:
- Wristbands with barcodes that have to be scanned each time the nurse visits their room
- An embedded RFID transmitter in a cut-resistant bracelet on the baby's leg that allows staff to see on a locational display where the baby is at any time and to sound an alarm if the infant is taken outside the protective area
These systems link the baby to the mother, but what actually documents the identity of the baby? The paper card with the baby's left and right footprints and the mother's right thumbprint has been used for decades, but is that sufficient for the future?
This issue of infant authentication reminded me of a presentation I recently attended given by noted educator and biometrics researcher Professor Anil Jain at Michigan State University. Jain and his team worked under a grant from the Bill and Melinda Gates Foundation to develop a reliable, low-cost authentication process for young children. The primary purpose was to enable the tracking of children's vaccination schedules to ensure that the right child receives the full regimen of immunizations. One of the critical issues Jain and his team faced is the difficulty in obtaining usable fingerprints from newborns—the skin on their fingertips is pliable, which results in poor contrast between the pattern of their ridges and valleys.
The goal of the research program was to determine the earliest possible age at which reliable fingerprints could be obtained using current technology. Using a high-resolution optical reader providing a fast capture rate (infants don't like to be still for very long), the research team found that fingerprint enrollment for children older than six months provides acceptance rates of 99 percent. This method can potentially serve as a reliable authentication method for the remainder of their life. Coupled with the creation of an electronic health registry, the health care worker needs only to scan a child's finger to bring up immunization records and determine any future vaccinations required. You can find a short presentation of Jain's work here.
While the public is likely to continue to question the overall benefits of biometrics, Jain's work shows an additional use for biometrics technology. Where else might biometric programs be applied?
By David Lott, a payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed