The use of passwords and pin numbers is part of our daily lives, being a necessity in ensuring our data and money doesn’t fall into the wrong hands. However passwords and pattern-based pins have their obvious limitations, and they are only as secure as the user is cautious. One method of improving security utilises biometric technology, which is based on the biological or behavioural characteristics of an individual. Biometric-based security systems are certainly nothing new. The concept of using fingerprints, retinal scans and voice recognition as security measures materialised decades ago, and such techniques are frequently used for authentication purposes. Despite these technological developments, ongoing research is attempting to develop more robust and secure methods of identification.
Researchers at the University of Albany are developing a unique new technique of biometric identification using only a person’s sweat. Human sweat, and all body fluids for that matter, contains a plethora of chemical compounds, ranging from small weight molecules to large proteins. These compounds originate from a variety of sources, with some resulting from endogenous metabolic processes within the body, and others being introduced through diet and environmental exposure. Metabolite levels can be affected by an endless array of factors, including sex, ethnicity, age and lifestyle. Interestingly, it is now known that the presence and amount of some of these compounds can vary greatly between different people, thus in theory unique metabolome profiles could be harnessed for identification purposes.
The compounds the technique will focus on is vital, as certain chemical levels can fluctuate wildly throughout the day depending on what we have eaten, for instance. However levels of certain chemicals have been found to be relatively stable or at least only vary gradually. In this research, Assistant Professor Jan Halámek and his team focused on using amino acid profiles of sweat to offer a unique means of authentication.
By first establishing which amino acids are present in a person’s skin secretions, a wearable device can then be constructed which will monitor the levels of these compounds. The device would initially require a kind of enrolment period, during which time the user’s skin secretions would be constantly measured in order to develop a unique profile of metabolites. It is already known that the metabolites released by the body vary throughout the day, so such a monitoring period would be necessary to take into account these changes.
Over time a profile of the user’s skin secretions would be built up and stored within the device, acting as a kind of standard for comparison. When future skin secretions are analysed by the device, the profiles will be compared with the known user profile and used to confirm the identity of the user. In the event of anyone else picking up the device, the instrument would detect a different skin secretion profile and lock the device or turn it off, thus ensuring security of the smartphone or computer.
If successful, the technology could offer an improved active authentication system, either as a standalone system or supplementing existing technology. However the technique is very much in its infancy and a great deal more research will be required before this kind of technology is rolled out commercially, if it ever is possible. It is likely that such a technique will be affected by contamination, for instance as the user’s hands become dirty throughout the day or if cleaning or cosmetic products are applied to the skin. Furthermore, if authentication is based on comparison with an electronically stored profile, the device may still be susceptible to hacking in order to bypass the security system. But if this technique could reach a sufficient level of robustness, the days of struggling to remember your password could be eliminated.
Agudelo, J. Privman, V. Halamek, J. Promises and Challenges in Continuous Tracking Utilizing Amino Acids in Skin Secretions for Active Multi-Factor Biometric Authentication for Cybersecurity. ChemPhysChem. 18, 1714-1720 (2017).