NRC Report: Biometric Technology “Inherently Fallible”

Biometric systems — designed to recognize individuals based on biological and behavioral traits such as fingerprints, palm prints, or voice or face recognition — are “inherently fallible,” says a new report by the National Research Council (NRC). In addition, says the report, no single biological and behavioral trait has been identified that is stable and distinctive across all groups. To strengthen the science and improve system effectiveness, additional research is needed at virtually all levels of design and operation.

The NRC report, Biometric Recognition: Challenges and Opportunities, was edited by Joseph N. Pato and Lynette I. Millett. Pato, a technologist at Hewlett-Packard’s HP Laboratories in Palo Alto, California says in the  report’s announcement, “While some biometric systems can be effective for specific tasks, they are not nearly as infallible as their depiction in popular culture might suggest.” Pato adds, “Bolstering the science is essential to gain a complete understanding of the strengths and limitations of these systems.”

Biometric systems are increasingly used to regulate access to facilities, information, and other rights or benefits, but questions persist about their effectiveness as security or surveillance mechanisms. The systems provide “probabilistic results,” which means there’s an inherent uncertainty in any given system, the report says. It notes that when the likelihood of an impostor is rare, even systems with very accurate sensors and matching capabilities can have a high false-alarm rate. This could become costly or even dangerous in systems designed to provide heightened security.

The report identifies numerous sources of uncertainty in the systems that need to be considered in system design and operation. For example, biometric characteristics may vary over an individual’s lifetime due to age, stress, disease, or other factors. Technical issues regarding calibration of sensors, degradation of data, and security breaches also contribute to variability in these systems.

Biometric systems need to be designed and evaluated against their specific intended purposes and the contexts in which they are being used, the report says. The report also identifies several features that a biometric system should contain. Systems should be designed to anticipate and plan for errors, even if they are expected to be infrequent.

The report calls for more research in all aspects of design and operation, from studying the distribution of biometric traits in given populations to understanding how people interact with the technologies.

Photo: Glenn J. Mason/Flickr